Predicting the SPY's Future Closing Price with a Multi-Model Forecast

Creating many machine learning models to predict future price movements from Redis.

How?

  1. Uses pricing metrics (hlocv)
  2. Streamline development and deployment of machine learning forecasts by storing large, pre-trained models living in Redis
  3. Custom rolled dataset (takes about 7 hours per 1 ticker)
  4. Technical indicators

Why?

  1. Took too long to manually rebuild the dataset, and build + tune new models
  2. Improve model accuracy by tracking success (situational/seasonal risks)
  3. Wanted simple, consistent delivery of results
  4. Service layer for abstracting model implementation
  5. Multi-tenant, distributed machine learning cloud
  6. Team needed Jupyter integration
  7. Data security - so it had to run on-premise and cloud

Now it takes 30 minutes to build the dataset and 5 minutes to make new predictions

Sample SPY Multi-Model Forecast

Setup the Environment

Load the shared core, methods, and environment before starting processing


In [1]:
from __future__ import print_function
import sys, os, requests, json, datetime

# Load the environment and login the user
from src.common.load_redten_ipython_env import user_token, user_login, csv_file, run_job, core, api_urls, ppj, rt_url, rt_user, rt_pass, rt_email, lg, good, boom, anmt, mark, ppj, uni_key, rest_login_as_user, rest_full_login, wait_for_job_to_finish, wait_on_job, get_job_analysis, get_job_results, get_analysis_manifest, get_job_cache_manifest, build_prediction_results, build_forecast_results, get_job_cache_manifest, search_ml_jobs, show_logs, show_errors, ipyImage, ipyHTML, ipyDisplay, pd, np

Configure the job


In [2]:
# dataset name is the ticker
ds_name = "SPY"

# Label and description for job
title = str(ds_name) + " Forecast v5 - " + str(uni_key())
desc = "Forecast simulation - " + str(datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S"))

# Whats the algorithm model you want to use?
algo_name = "xgb-regressor"

# If your dataset is stored in redis, you can pass in the location
# to the dataset like: <redis endpoint>:<key>
rloc = "" 

# If your dataset is stored in S3, you can pass in the location 
# to the dataset like: <bucket>:<key>
sloc = "" 

# During training what ratio of tests-vs-training do you want to use?
# Trade off smarts vs accuracy...how smart are we going?
test_ratio = 0.1

# Customize dataset samples used during the analysis using json dsl
sample_filter_rules = {}

# What column do you want to predict values?
target_column_name = "FClose"

# What columns can the algorithms use for training and learning?
feature_column_names = [ "FHigh", "FLow", "FOpen", "FClose", "FVolume" ]

# values in the Target Column
target_column_values = [ "GoodBuys", "BadBuys", "Not Finished" ] 

# How many units ahead do you want to forecast?
units_ahead_set = [ 5, 10, 15, 20, 25, 30 ]
units_ahead_type = "Days"

# Prune non-int/float columns as needed: 
ignore_features = [       
                "Ticker",
                "Date",
                "FDate",
                "FPrice",
                "DcsnDate",
                "Decision"
            ]

# Set up the XGB parameter
# https://github.com/dmlc/xgboost/blob/master/doc/parameter.md
train_xgb = {
                "learning_rate" : 0.20, 
                "num_estimators" : 50, 
                "sub_sample" : 0.20, 
                "col_sample_by_tree" : 0.90, 
                "col_sample_by_level" : 1.0, 
                "objective" : "reg:linear",
                "max_depth" : 3,
                "max_delta_step" : 0,
                "min_child_weight" : 1, 
                "reg_alpha" : 0, 
                "reg_lambda" : 1,
                "base_score" : 0.6,
                "gamma" : 0,
                "seed" : 42, 
                "silent" : True
            } 

# Predict new price points during the day
predict_row = {
                "High"   : 250.82,
                "Low"    : 245.54,
                "Open"   : 247.77,
                "Close"  : 246.24,
                "Volume" : 77670266
            }

Start Forecasting


In [3]:
job_id = None # on success, this will store the actively running job's id
csv_file = ""
post_data = {
            "predict_this_data" : predict_row,
            "title" : title,
            "desc" : desc,
            "ds_name" : ds_name,
            "target_column_name" : target_column_name,
            "feature_column_names" : feature_column_names,
            "ignore_features" : ignore_features,
            "csv_file" : csv_file,
            "rloc" : rloc,
            "sloc" : sloc,
            "algo_name" : algo_name,
            "test_ratio" : test_ratio,
            "target_column_values" : target_column_values,
            "label_column_name" : target_column_name, 
            "prediction_type" : "Forecast",
            "ml_type" : "Playbook-UnitsAhead",
            "train" : train_xgb,
            "tracking_type" : "",
            "units_ahead_set" : units_ahead_set,
            "units_ahead_type" : units_ahead_type,
            "forecast_type" : "ETFPriceForecasting",
            "sample_filters" : sample_filter_rules,
            "predict_units_back" : 90, # how many days back should the final chart go?
            "send_to_email" : [ "jay.p.h.johnson@gmail.com" ] # comma separated list
        }

anmt("Running job: " + str(title))

auth_headers = { 
            "Content-type": "application/json",
            "Authorization" : "JWT " + str(user_token)
        }

job_response = run_job(post_data=post_data, headers=auth_headers)

if job_response["status"] != "valid":
    boom("Forecast job failed with error=" + str(job_response["status"]))
else:
    if "id" not in job_response["data"]:
        boom("Failed to create new forecast job")
    else:
        job_id = job_response["data"]["id"]
        job_status = job_response["data"]["status"]
        lg("Started Forecast job=" + str(job_id) + " with current status=" + str(job_status))
# end of if job was valid or not


Running job: SPY Forecast v5 - 34d6bf175cad4aaf9234aa06e95f4e4
Started Forecast job=558 with current status=requested

Wait for the job to finish


In [4]:
job_data = {}
job_report = {}

# Should hook this up to a randomized image loader...
ipyDisplay(ipyImage(url="https://media.giphy.com/media/l397998l2DT0ogare/giphy.gif"))

job_res = {}
if job_id == None:
    boom("Failed to start a new job")
else:
    job_res = wait_on_job(job_id)

    if job_res["status"] != "SUCCESS":
        boom("Job=" + str(job_id) + " failed with status=" + str(job_res["status"]) + " err=" + str(job_res["error"]))
    else:
        job_data = job_res["record"]
        anmt("Job Report:")
        lg(ppj(job_data), 5)
# end of waiting


Waiting on job=558 url=https://redten.io/ml/558/
Job=558 is training - Step 3/10
Job=558 is analyzing - Step 5/10
Job=558 is caching - Step 6/10
Job=558 is plotting - Step 7/10
Job=558 is uploading - Step 9/10
Job=558 completed
Job Report:
{
    "job": {
        "algo_name": "xgb-regressor",
        "control_state": "active",
        "created": "2017-05-26 08-02-13",
        "csv_file": "",
        "desc": "Forecast simulation - 2017-05-26 08:02:13",
        "ds_name": "SPY",
        "feature_column_names": [
            "FHigh",
            "FLow",
            "FOpen",
            "FClose",
            "FVolume"
        ],
        "id": 558,
        "ignore_features": [
            "Ticker",
            "Date",
            "FDate",
            "FPrice",
            "DcsnDate",
            "Decision"
        ],
        "images": [
            {
                "author_name": null,
                "desc": null,
                "id": 12122,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12122_7c49dccd964548fc.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 5-Days - Predictive Accuracy\nPredicted Close 5 Days vs Actual Close 5 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12121,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12121_5ae84fdafa034174.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 5-Days - Predictive Accuracy\nPredicted High 5 Days vs Actual High 5 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12120,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12120_2d87f32772344921.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 10-Days - Predictive Accuracy\nPredicted Close 10 Days vs Actual Close 10 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12119,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12119_c54a1003037d4eb3.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 15-Days - Predictive Accuracy\nPredicted Close 15 Days vs Actual Close 15 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12118,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12118_192e7a3c4ba840fc.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 30-Days - Predictive Accuracy\nPredicted Close 30 Days vs Actual Close 30 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12117,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12117_6b74cc14075f4f20.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 25-Days - Predictive Accuracy\nPredicted Low 25 Days vs Actual Low 25 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12116,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12116_1fed112e3a4e42ff.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 20-Days - Predictive Accuracy\nPredicted Low 20 Days vs Actual Low 20 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12115,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12115_795d9657c2534f76.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 15-Days - Predictive Accuracy\nPredicted Volume 15 Days vs Actual Volume 15 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12114,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12114_c19203024ec84944.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 30-Days - Predictive Accuracy\nPredicted Volume 30 Days vs Actual Volume 30 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12113,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12113_1b000088cf154451.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 10-Days - Predictive Accuracy\nPredicted Volume 10 Days vs Actual Volume 10 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12112,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12112_c72e079979774d17.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 20-Days - Predictive Accuracy\nPredicted Open 20 Days vs Actual Open 20 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12111,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12111_b0c21c4c4d0d4e7a.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 30-Days - Predictive Accuracy\nPredicted High 30 Days vs Actual High 30 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12110,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12110_2a27883839464a6c.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 25-Days - Predictive Accuracy\nPredicted Open 25 Days vs Actual Open 25 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12109,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12109_325e5fefa07a444d.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 15-Days - Predictive Accuracy\nPredicted High 15 Days vs Actual High 15 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12108,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12108_7dfe5cad6dd341c2.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 10-Days - Predictive Accuracy\nPredicted High 10 Days vs Actual High 10 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12107,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12107_2fcaa612a10346d9.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 15-Days - Predictive Accuracy\nPredicted Low 15 Days vs Actual Low 15 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12106,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12106_29c5e186270348f4.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 5-Days - Predictive Accuracy\nPredicted Open 5 Days vs Actual Open 5 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12105,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12105_7139a2dd4df5447b.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 10-Days - Predictive Accuracy\nPredicted Low 10 Days vs Actual Low 10 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12104,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12104_74d4b72e2bd342d6.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 25-Days - Predictive Accuracy\nPredicted Close 25 Days vs Actual Close 25 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12103,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12103_286431b3eabc4af5.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 5-Days - Predictive Accuracy\nPredicted Volume 5 Days vs Actual Volume 5 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12102,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12102_b8f6174000f5455c.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 20-Days - Predictive Accuracy\nPredicted Close 20 Days vs Actual Close 20 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12101,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12101_bfe9888ad15648f9.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 30-Days - Predictive Accuracy\nPredicted Low 30 Days vs Actual Low 30 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12100,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12100_0761ea1b0f684ce9.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 25-Days - Predictive Accuracy\nPredicted Volume 25 Days vs Actual Volume 25 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12099,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12099_8f73f41290794595.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 20-Days - Predictive Accuracy\nPredicted Volume 20 Days vs Actual Volume 20 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12098,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12098_5fd3e28826754f24.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 20-Days - Predictive Accuracy\nPredicted High 20 Days vs Actual High 20 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12097,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12097_214858c79e704f4c.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 30-Days - Predictive Accuracy\nPredicted Open 30 Days vs Actual Open 30 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12096,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12096_168bcf41ff5142ad.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 25-Days - Predictive Accuracy\nPredicted High 25 Days vs Actual High 25 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12095,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12095_de06fd5548504b41.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 15-Days - Predictive Accuracy\nPredicted Open 15 Days vs Actual Open 15 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12094,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12094_354bb5267c694ba5.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 5-Days - Predictive Accuracy\nPredicted Low 5 Days vs Actual Low 5 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12093,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12093_e2ecdc03bc8a4f30.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY-2-558 10-Days - Predictive Accuracy\nPredicted Open 10 Days vs Actual Open 10 Days",
                "version": 1
            },
            {
                "author_name": null,
                "desc": null,
                "id": 12092,
                "image": "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12092_954ad32f51ef47f4.png",
                "json_data": null,
                "label": null,
                "shareable_link": null,
                "sloc": null,
                "status": "initial",
                "title": "SPY Close forecast overlay between 2017-02-27 00:00:00 - 2017-05-15 00:00:00",
                "version": 1
            }
        ],
        "manifest_secret_key": "live_2_f3f31f711f1b4081862a1927be0758880c3b3f00e2848aeb2a7f68997b4e17",
        "manifest_sloc": "redten-models-west:SPY-2-558_manifest.json",
        "max_features": 10,
        "ml_type": "Playbook-UnitsAhead",
        "prediction_type": "forecast",
        "rloc": "",
        "sloc": "",
        "status": "completed",
        "target_column_name": "FClose",
        "target_column_values": [
            "GoodBuys",
            "BadBuys",
            "Not Finished"
        ],
        "title": "SPY Forecast v5 - 34d6bf175cad4aaf9234aa06e95f4e4",
        "units_ahead_set": [
            5,
            10,
            15,
            20,
            25,
            30
        ],
        "units_ahead_type": "Days",
        "updated": "2017-05-26 08-03-53",
        "version": 1
    }
}

Get Forecast Accuracies


In [5]:
job_report = {}
if job_id == None:
    boom("Failed to start a new job")
else:
    # Get the analysis, but do not auto-show the plots
    job_report = get_job_analysis(job_id, show_plots=False)
    if len(job_report) == 0:
        boom("Job=" + str(job_id) + " failed")
    else:
        lg("")
    # if the job failed
# end of get job analysis

# Build the forecast accuracy dictionary from the analysis
# and show the forecast dataframes
acc_results = build_forecast_results(job_report)
for col in acc_results:
    col_node = acc_results[col]
    
    predictions_df = col_node["predictions_df"]
    date_predictions_df = col_node["date_predictions_df"]
    train_predictions_df = col_node["train_predictions_df"]
    
    lg("--------------------------------------------------")
    # for all columns in the accuracy dictionary:  
    # successful predictions above 90%...how's that error rate though?
    if col_node["accuracy"] > 0.90: 
        good("Column=" + str(col) + " accuracy=" + str(col_node["accuracy"]) + " mse=" + str(col_node["mse"]) + " num_predictions=" + str(len(col_node["date_predictions_df"].index)))
    # successful predictions between 90% and 80%...how's that error rate though?
    elif 0.90 > col_node["accuracy"] > 0.80:
        lg("Column=" + str(col) + " accuracy=" + str(col_node["accuracy"]) + " mse=" + str(col_node["mse"]) + " num_predictions=" + str(len(col_node["date_predictions_df"].index)))
    else:
        boom("Column=" + str(col) + " is not very accurate: accuracy=" + str(col_node["accuracy"]) + " mse=" + str(col_node["mse"]) + " num_predictions=" + str(len(col_node["predictions_df"].index)))
    # end of header line
    
    # show the timeseries forecast
    ipyDisplay(date_predictions_df)
    
    lg("")
# end of showing prediction results


Getting analysis for job=558 url=https://redten.io/ml/analysis/558/
SUCCESS - GET Analysis Response Status=200 Reason=OK
Found Job=558 analysis

--------------------------------------------------
Column=FOpen_10 accuracy=0.99848923125 mse=1.87511875363 num_predictions=50
COpen Date FOpen
0 236.64 2017-02-27 237.073318
1 236.67 2017-02-28 236.157669
2 238.39 2017-03-01 237.216782
3 239.56 2017-03-02 235.210175
4 238.17 2017-03-03 233.699661
5 237.50 2017-03-06 237.479095
6 237.71 2017-03-07 235.920334
7 237.34 2017-03-08 237.982605
8 236.70 2017-03-09 237.568695
9 237.97 2017-03-10 235.905457
10 237.62 2017-03-13 236.509735
11 237.18 2017-03-14 234.706680
12 237.56 2017-03-15 236.196335
13 239.11 2017-03-16 235.618851
14 237.75 2017-03-17 236.832428
15 237.03 2017-03-20 236.587128
16 237.47 2017-03-21 237.348419
17 233.77 2017-03-22 237.528641
18 234.00 2017-03-23 234.676941
19 234.38 2017-03-24 236.232239
20 231.93 2017-03-27 235.156250
21 233.27 2017-03-28 234.025803
22 234.99 2017-03-29 234.184830
23 235.47 2017-03-30 233.101852
24 235.90 2017-03-31 237.531754
25 235.80 2017-04-03 233.742493
26 234.95 2017-04-04 235.229843
27 236.26 2017-04-05 236.990433
28 234.94 2017-04-06 237.611145
29 235.15 2017-04-07 234.539597
30 235.36 2017-04-10 237.734253
31 234.90 2017-04-11 236.622452
32 234.74 2017-04-12 234.549881
33 233.64 2017-04-13 236.684601
34 233.64 2017-04-14 237.827515
35 233.11 2017-04-17 235.516617
36 233.72 2017-04-18 236.572464
37 234.52 2017-04-19 237.910950
38 235.25 2017-04-21 237.734253
39 237.18 2017-04-24 233.342529
40 237.91 2017-04-25 237.520111
41 238.51 2017-04-26 232.563446
42 238.77 2017-04-27 234.030930
43 238.90 2017-04-28 234.826447
44 238.68 2017-05-01 236.066803
45 238.84 2017-05-02 235.076416
46 238.29 2017-05-03 233.416733
47 238.83 2017-05-04 236.058029
48 239.19 2017-05-05 235.863937
49 239.75 2017-05-08 235.429077
--------------------------------------------------
Column=FLow_5 accuracy=0.998086039927 mse=1.85004769663 num_predictions=55
CLow Date FLow
0 236.35 2017-02-27 236.929581
1 236.02 2017-02-28 236.010742
2 238.37 2017-03-01 233.294571
3 238.21 2017-03-02 232.308624
4 237.73 2017-03-03 234.968628
5 237.01 2017-03-06 236.273605
6 237.71 2017-03-07 236.508759
7 236.40 2017-03-08 233.406143
8 235.74 2017-03-09 235.658737
9 236.59 2017-03-10 231.818314
10 237.24 2017-03-13 234.524475
11 236.19 2017-03-14 235.794052
12 237.29 2017-03-15 233.459610
13 238.10 2017-03-16 233.198196
14 237.03 2017-03-17 237.028275
15 236.32 2017-03-20 233.921158
16 233.58 2017-03-21 235.768692
17 233.05 2017-03-22 236.422226
18 233.60 2017-03-23 236.376251
19 232.96 2017-03-24 233.915695
20 231.61 2017-03-27 236.468430
21 233.14 2017-03-28 236.915192
22 234.72 2017-03-29 235.886642
23 235.27 2017-03-30 233.052322
24 235.68 2017-03-31 235.961349
25 233.91 2017-04-03 232.261780
26 234.56 2017-04-04 234.373062
27 234.54 2017-04-05 233.631317
28 234.42 2017-04-06 234.610016
29 234.64 2017-04-07 236.890594
30 234.73 2017-04-10 234.642807
31 233.34 2017-04-11 234.695312
32 233.77 2017-04-12 233.865128
33 232.51 2017-04-13 236.446716
34 232.51 2017-04-14 235.104309
35 232.88 2017-04-17 233.097855
36 233.08 2017-04-18 236.518036
37 233.18 2017-04-19 236.571960
38 234.13 2017-04-21 234.559448
39 234.56 2017-04-24 236.496582
40 237.81 2017-04-25 236.412048
41 238.35 2017-04-26 236.186478
42 237.98 2017-04-27 235.107819
43 237.93 2017-04-28 233.527512
44 238.20 2017-05-01 236.808197
45 238.30 2017-05-02 236.269989
46 237.70 2017-05-03 235.095825
47 237.78 2017-05-04 234.935760
48 238.68 2017-05-05 235.331024
49 239.17 2017-05-08 234.041565
50 239.04 2017-05-09 236.181931
51 239.04 2017-05-10 236.355286
52 238.13 2017-05-11 232.871155
53 238.67 2017-05-12 233.601791
54 239.45 2017-05-15 236.600067
--------------------------------------------------
Column=FOpen_15 accuracy=0.998673216637 mse=1.36712346617 num_predictions=45
COpen Date FOpen
0 236.64 2017-02-27 238.666595
1 236.67 2017-02-28 237.613739
2 238.39 2017-03-01 237.458069
3 239.56 2017-03-02 236.777634
4 238.17 2017-03-03 238.085297
5 237.50 2017-03-06 237.372849
6 237.71 2017-03-07 234.226608
7 237.34 2017-03-08 235.627625
8 236.70 2017-03-09 236.552750
9 237.97 2017-03-10 234.387222
10 237.62 2017-03-13 234.385269
11 237.18 2017-03-14 234.300766
12 237.56 2017-03-15 238.415405
13 239.11 2017-03-16 234.556824
14 237.75 2017-03-17 235.208084
15 237.03 2017-03-20 235.048111
16 237.47 2017-03-21 236.497421
17 233.77 2017-03-22 234.135391
18 234.00 2017-03-23 234.963669
19 234.38 2017-03-24 234.820038
20 231.93 2017-03-27 237.683395
21 233.27 2017-03-28 238.560837
22 234.99 2017-03-29 238.095673
23 235.47 2017-03-30 236.939590
24 235.90 2017-03-31 237.795105
25 235.80 2017-04-03 237.874680
26 234.95 2017-04-04 233.943573
27 236.26 2017-04-05 234.943817
28 234.94 2017-04-06 238.051514
29 235.15 2017-04-07 235.384384
30 235.36 2017-04-10 237.187256
31 234.90 2017-04-11 234.732925
32 234.74 2017-04-12 233.787415
33 233.64 2017-04-13 238.512772
34 233.64 2017-04-14 234.131561
35 233.11 2017-04-17 238.079544
36 233.72 2017-04-18 233.760864
37 234.52 2017-04-19 234.805496
38 235.25 2017-04-21 234.080505
39 237.18 2017-04-24 236.442703
40 237.91 2017-04-25 233.432205
41 238.51 2017-04-26 234.990753
42 238.77 2017-04-27 236.403168
43 238.90 2017-04-28 235.337692
44 238.68 2017-05-01 237.205765
--------------------------------------------------
Column=FHigh_25 accuracy=0.99852625367 mse=2.07955530003 num_predictions=35
CHigh Date FHigh
0 237.31 2017-02-27 238.076721
1 236.95 2017-02-28 234.146927
2 240.32 2017-03-01 237.973175
3 239.57 2017-03-02 237.950897
4 238.61 2017-03-03 237.457062
5 238.12 2017-03-06 238.205643
6 237.71 2017-03-07 237.868561
7 237.64 2017-03-08 233.265198
8 237.24 2017-03-09 234.387634
9 238.02 2017-03-10 237.551636
10 237.86 2017-03-13 235.177246
11 237.24 2017-03-14 237.380798
12 239.44 2017-03-15 234.876892
13 239.20 2017-03-16 235.736435
14 237.97 2017-03-17 238.128571
15 237.36 2017-03-20 237.794815
16 237.61 2017-03-21 237.752106
17 234.61 2017-03-22 236.209839
18 235.34 2017-03-23 234.751801
19 235.04 2017-03-24 234.983871
20 233.92 2017-03-27 238.090561
21 235.81 2017-03-28 233.964142
22 235.81 2017-03-29 237.529388
23 236.52 2017-03-30 234.443344
24 236.51 2017-03-31 238.910355
25 236.03 2017-04-03 237.520447
26 235.58 2017-04-04 237.694489
27 237.39 2017-04-05 234.321915
28 236.00 2017-04-07 238.316208
29 236.26 2017-04-10 237.038849
30 235.18 2017-04-11 237.573196
31 234.96 2017-04-12 235.150558
32 234.49 2017-04-13 238.119049
33 234.49 2017-04-14 234.923492
34 234.57 2017-04-17 235.094742
--------------------------------------------------
Column=FOpen_30 accuracy=0.998227319751 mse=3.31048708482 num_predictions=30
COpen Date FOpen
0 236.64 2017-02-27 234.693985
1 236.67 2017-02-28 236.030411
2 238.39 2017-03-01 235.485245
3 239.56 2017-03-02 237.560745
4 238.17 2017-03-03 235.975784
5 237.50 2017-03-06 235.504822
6 237.34 2017-03-08 235.342697
7 236.70 2017-03-09 236.801010
8 237.97 2017-03-10 237.499802
9 237.62 2017-03-13 234.917587
10 237.18 2017-03-14 233.616501
11 237.56 2017-03-15 237.476257
12 239.11 2017-03-16 232.399368
13 237.75 2017-03-17 236.711685
14 237.03 2017-03-20 236.574722
15 237.47 2017-03-21 237.442719
16 233.77 2017-03-22 234.120331
17 234.00 2017-03-23 233.750183
18 234.38 2017-03-24 237.016220
19 231.93 2017-03-27 232.151428
20 233.27 2017-03-28 236.966064
21 234.99 2017-03-29 237.586349
22 235.47 2017-03-30 236.353043
23 235.90 2017-03-31 237.411346
24 235.80 2017-04-03 238.304245
25 234.95 2017-04-04 235.202972
26 236.26 2017-04-05 236.926498
27 234.94 2017-04-06 237.732910
28 235.15 2017-04-07 236.833130
29 235.36 2017-04-10 238.301956
--------------------------------------------------
Column=FHigh_20 accuracy=0.99830741251 mse=1.25659633163 num_predictions=40
CHigh Date FHigh
0 237.31 2017-02-27 237.670395
1 236.95 2017-02-28 235.000000
2 240.32 2017-03-01 235.114883
3 239.57 2017-03-02 238.374283
4 238.61 2017-03-03 236.344513
5 238.12 2017-03-06 237.315735
6 237.71 2017-03-07 236.214111
7 237.64 2017-03-08 237.246307
8 237.24 2017-03-09 238.396393
9 238.02 2017-03-10 236.479446
10 237.86 2017-03-13 237.763672
11 237.24 2017-03-14 235.117249
12 239.44 2017-03-15 234.345337
13 239.20 2017-03-16 236.225082
14 237.97 2017-03-17 234.697128
15 237.36 2017-03-20 238.262817
16 237.61 2017-03-21 235.133362
17 234.61 2017-03-22 238.553192
18 235.34 2017-03-23 238.866257
19 235.04 2017-03-24 233.853409
20 233.92 2017-03-27 236.915115
21 235.81 2017-03-28 236.299088
22 235.81 2017-03-29 234.427261
23 236.52 2017-03-30 235.588684
24 236.51 2017-03-31 237.638855
25 236.03 2017-04-03 238.629776
26 235.58 2017-04-04 234.676880
27 237.39 2017-04-05 238.158661
28 236.04 2017-04-06 238.226105
29 236.00 2017-04-07 235.759537
30 236.26 2017-04-10 238.553192
31 235.18 2017-04-11 237.456482
32 234.96 2017-04-12 238.249573
33 234.49 2017-04-13 235.751602
34 234.49 2017-04-14 237.794708
35 234.57 2017-04-17 235.746872
36 234.49 2017-04-18 237.754822
37 234.95 2017-04-19 236.131210
38 235.31 2017-04-21 235.008286
39 237.41 2017-04-24 238.001404
--------------------------------------------------
Column=FVolume_20 accuracy=0.908394205285 mse=1.16457020726e+14 num_predictions=40
CVolume Date FVolume
0 56515440.0 2017-02-27 65846012.0
1 96961938.0 2017-02-28 83015680.0
2 149158170.0 2017-03-01 64974404.0
3 70245978.0 2017-03-02 68879856.0
4 81974300.0 2017-03-03 66677188.0
5 55391533.0 2017-03-06 101378352.0
6 393822.0 2017-03-07 99078096.0
7 78168795.0 2017-03-08 97593016.0
8 90683918.0 2017-03-09 64936852.0
9 81991652.0 2017-03-10 95077096.0
10 57256824.0 2017-03-13 61924532.0
11 59880778.0 2017-03-14 83244072.0
12 96081750.0 2017-03-15 76859560.0
13 78343951.0 2017-03-16 63056964.0
14 89002111.0 2017-03-17 62771728.0
15 52536979.0 2017-03-20 62433760.0
16 131809275.0 2017-03-21 79715272.0
17 97569204.0 2017-03-22 57779620.0
18 100410277.0 2017-03-23 67151176.0
19 112504853.0 2017-03-24 89839216.0
20 87454452.0 2017-03-27 81388680.0
21 93483915.0 2017-03-28 84004696.0
22 61950354.0 2017-03-29 89794744.0
23 56737890.0 2017-03-30 87538128.0
24 73733094.0 2017-03-31 57832536.0
25 85546486.0 2017-04-03 77662536.0
26 56466195.0 2017-04-04 80301888.0
27 108800604.0 2017-04-05 64505432.0
28 69135757.0 2017-04-06 155792864.0
29 74412311.0 2017-04-07 59956728.0
30 67615302.0 2017-04-10 63884772.0
31 88045276.0 2017-04-11 61656740.0
32 81864436.0 2017-04-12 46140620.0
33 92880394.0 2017-04-13 72249024.0
34 92880394.0 2017-04-14 83007760.0
35 68405367.0 2017-04-17 83554112.0
36 83225821.0 2017-04-18 107252736.0
37 68699868.0 2017-04-19 107173648.0
38 110389847.0 2017-04-21 74443776.0
39 119209877.0 2017-04-24 55639312.0
--------------------------------------------------
Column=FVolume_25 accuracy=0.875114335535 mse=5.61678331942e+13 num_predictions=35
CVolume Date FVolume
0 56515440.0 2017-02-27 109613136.0
1 96961938.0 2017-02-28 84890608.0
2 149158170.0 2017-03-01 56550484.0
3 70245978.0 2017-03-02 64852992.0
4 81974300.0 2017-03-03 78665056.0
5 55391533.0 2017-03-06 62343776.0
6 393822.0 2017-03-07 65068232.0
7 78168795.0 2017-03-08 73619968.0
8 90683918.0 2017-03-09 86562520.0
9 81991652.0 2017-03-10 83784728.0
10 57256824.0 2017-03-13 70358608.0
11 59880778.0 2017-03-14 94782192.0
12 96081750.0 2017-03-15 90527832.0
13 78343951.0 2017-03-16 82991488.0
14 89002111.0 2017-03-17 58278716.0
15 52536979.0 2017-03-20 112854176.0
16 131809275.0 2017-03-21 55025048.0
17 97569204.0 2017-03-22 66197604.0
18 100410277.0 2017-03-23 75862600.0
19 112504853.0 2017-03-24 65388524.0
20 87454452.0 2017-03-27 67723848.0
21 93483915.0 2017-03-28 113617928.0
22 61950354.0 2017-03-29 94782192.0
23 56737890.0 2017-03-30 62635744.0
24 73733094.0 2017-03-31 164215552.0
25 85546486.0 2017-04-03 60994780.0
26 56466195.0 2017-04-04 57976436.0
27 108800604.0 2017-04-05 60855060.0
28 74412311.0 2017-04-07 61561904.0
29 67615302.0 2017-04-10 60636220.0
30 88045276.0 2017-04-11 66711616.0
31 81864436.0 2017-04-12 99195736.0
32 92880394.0 2017-04-13 68778016.0
33 92880394.0 2017-04-14 74303160.0
34 68405367.0 2017-04-17 109592048.0
--------------------------------------------------
Column=FLow_30 accuracy=0.998277970848 mse=2.96286991645 num_predictions=30
CLow Date FLow
0 236.35 2017-02-27 235.590820
1 236.02 2017-02-28 236.062424
2 238.37 2017-03-01 235.890472
3 238.21 2017-03-02 236.847061
4 237.73 2017-03-03 234.398666
5 237.01 2017-03-06 234.975235
6 236.40 2017-03-08 234.611038
7 235.74 2017-03-09 235.617630
8 236.59 2017-03-10 236.238098
9 237.24 2017-03-13 233.991272
10 236.19 2017-03-14 233.256729
11 237.29 2017-03-15 236.445068
12 238.10 2017-03-16 233.417892
13 237.03 2017-03-17 235.901810
14 236.32 2017-03-20 236.761841
15 233.58 2017-03-21 237.303406
16 233.05 2017-03-22 233.292572
17 233.60 2017-03-23 231.795105
18 232.96 2017-03-24 236.326782
19 231.61 2017-03-27 233.716980
20 233.14 2017-03-28 237.041916
21 234.72 2017-03-29 236.383987
22 235.27 2017-03-30 237.199799
23 235.68 2017-03-31 235.502136
24 233.91 2017-04-03 236.134262
25 234.56 2017-04-04 233.488770
26 234.54 2017-04-05 235.824600
27 234.42 2017-04-06 236.688004
28 234.64 2017-04-07 234.916611
29 234.73 2017-04-10 236.598770
--------------------------------------------------
Column=FClose_20 accuracy=0.998412598936 mse=1.17535948971 num_predictions=40
CClose Date FClose
0 237.11 2017-02-27 237.442490
1 236.47 2017-02-28 232.841202
2 239.78 2017-03-01 235.384415
3 238.27 2017-03-02 237.826065
4 238.42 2017-03-03 236.225693
5 237.71 2017-03-06 237.156509
6 237.71 2017-03-07 236.379303
7 236.56 2017-03-08 236.556427
8 236.86 2017-03-09 238.205902
9 237.69 2017-03-10 234.902100
10 237.81 2017-03-13 237.537613
11 236.90 2017-03-14 234.018509
12 238.95 2017-03-15 234.722977
13 238.48 2017-03-16 235.003632
14 237.03 2017-03-17 234.271652
15 236.77 2017-03-20 237.974792
16 233.73 2017-03-21 234.114609
17 234.28 2017-03-22 238.024323
18 234.03 2017-03-23 237.974792
19 233.86 2017-03-24 233.514282
20 233.62 2017-03-27 237.471954
21 235.32 2017-03-28 235.134399
22 235.54 2017-03-29 234.860138
23 236.29 2017-03-30 234.930237
24 235.74 2017-03-31 237.970367
25 235.33 2017-04-03 237.838745
26 235.48 2017-04-04 234.534256
27 234.78 2017-04-05 237.864243
28 235.44 2017-04-06 238.316711
29 235.20 2017-04-07 235.626526
30 235.34 2017-04-10 238.132874
31 235.06 2017-04-11 237.999649
32 234.03 2017-04-12 237.994370
33 232.51 2017-04-13 235.509064
34 234.03 2017-04-14 237.547440
35 234.57 2017-04-17 234.448135
36 233.87 2017-04-18 237.110519
37 233.44 2017-04-19 236.106888
38 234.59 2017-04-21 234.382721
39 237.17 2017-04-24 237.820465
--------------------------------------------------
Column=FVolume_5 accuracy=0.860508546731 mse=1.54519726446e+14 num_predictions=55
CVolume Date FVolume
0 56515440.0 2017-02-27 47996392.0
1 96961938.0 2017-02-28 46576272.0
2 149158170.0 2017-03-01 84940464.0
3 70245978.0 2017-03-02 90956808.0
4 81974300.0 2017-03-03 74963216.0
5 55391533.0 2017-03-06 55247800.0
6 393822.0 2017-03-07 69506984.0
7 78168795.0 2017-03-08 78182704.0
8 90683918.0 2017-03-09 51209964.0
9 81991652.0 2017-03-10 72567152.0
10 57256824.0 2017-03-13 97008752.0
11 59880778.0 2017-03-14 72061368.0
12 96081750.0 2017-03-15 97580816.0
13 78343951.0 2017-03-16 62746008.0
14 89002111.0 2017-03-17 69350312.0
15 52536979.0 2017-03-20 62146480.0
16 131809275.0 2017-03-21 82602136.0
17 97569204.0 2017-03-22 60284568.0
18 100410277.0 2017-03-23 61597948.0
19 112504853.0 2017-03-24 76730536.0
20 87454452.0 2017-03-27 49964956.0
21 93483915.0 2017-03-28 47066652.0
22 61950354.0 2017-03-29 177215760.0
23 56737890.0 2017-03-30 89014920.0
24 73733094.0 2017-03-31 69948528.0
25 85546486.0 2017-04-03 69864920.0
26 56466195.0 2017-04-04 52336716.0
27 108800604.0 2017-04-05 74350704.0
28 69135757.0 2017-04-06 113147264.0
29 74412311.0 2017-04-07 44125816.0
30 67615302.0 2017-04-10 76967448.0
31 88045276.0 2017-04-11 109703480.0
32 81864436.0 2017-04-12 71968256.0
33 92880394.0 2017-04-13 55783700.0
34 92880394.0 2017-04-14 70136248.0
35 68405367.0 2017-04-17 76711440.0
36 83225821.0 2017-04-18 59971112.0
37 68699868.0 2017-04-19 9440932.0
38 110389847.0 2017-04-21 66461792.0
39 119209877.0 2017-04-24 67064808.0
40 76698265.0 2017-04-25 59714204.0
41 84702455.0 2017-04-26 67836168.0
42 57410326.0 2017-04-27 118460488.0
43 63532845.0 2017-04-28 53469000.0
44 66882521.0 2017-05-01 46206144.0
45 57375732.0 2017-05-02 42782120.0
46 73137731.0 2017-05-03 100944264.0
47 61462732.0 2017-05-04 76188488.0
48 62001269.0 2017-05-05 99223600.0
49 48385730.0 2017-05-08 77355240.0
50 51363200.0 2017-05-09 80711072.0
51 51363200.0 2017-05-10 63145664.0
52 23215447.0 2017-05-11 90556112.0
53 51877428.0 2017-05-12 77122408.0
54 60025663.0 2017-05-15 45219076.0
--------------------------------------------------
Column=FClose_25 accuracy=0.99845288618 mse=2.04919274668 num_predictions=35
CClose Date FClose
0 237.11 2017-02-27 236.552734
1 236.47 2017-02-28 235.050507
2 239.78 2017-03-01 237.163391
3 238.27 2017-03-02 237.538895
4 238.42 2017-03-03 236.731735
5 237.71 2017-03-06 237.555984
6 237.71 2017-03-07 237.320831
7 236.56 2017-03-08 234.358047
8 236.86 2017-03-09 233.667374
9 237.69 2017-03-10 237.382278
10 237.81 2017-03-13 234.445786
11 236.90 2017-03-14 235.993500
12 238.95 2017-03-15 234.081650
13 238.48 2017-03-16 235.814377
14 237.03 2017-03-17 237.346420
15 236.77 2017-03-20 235.385986
16 233.73 2017-03-21 237.190720
17 234.28 2017-03-22 234.789978
18 234.03 2017-03-23 233.903641
19 233.86 2017-03-24 235.736145
20 233.62 2017-03-27 237.013290
21 235.32 2017-03-28 235.373993
22 235.54 2017-03-29 236.336746
23 236.29 2017-03-30 235.648193
24 235.74 2017-03-31 236.748077
25 235.33 2017-04-03 236.493500
26 235.48 2017-04-04 237.717545
27 234.78 2017-04-05 233.553162
28 235.20 2017-04-07 236.987137
29 235.34 2017-04-10 236.417114
30 235.06 2017-04-11 237.215149
31 234.03 2017-04-12 235.193405
32 232.51 2017-04-13 238.073685
33 234.03 2017-04-14 235.692825
34 234.57 2017-04-17 234.849548
--------------------------------------------------
Column=FLow_10 accuracy=0.998567461387 mse=2.11792497821 num_predictions=50
CLow Date FLow
0 236.35 2017-02-27 237.413834
1 236.02 2017-02-28 237.225479
2 238.37 2017-03-01 237.645981
3 238.21 2017-03-02 234.386292
4 237.73 2017-03-03 233.403427
5 237.01 2017-03-06 237.240616
6 237.71 2017-03-07 235.407959
7 236.40 2017-03-08 236.065186
8 235.74 2017-03-09 236.531815
9 236.59 2017-03-10 235.872589
10 237.24 2017-03-13 235.920380
11 236.19 2017-03-14 232.937210
12 237.29 2017-03-15 233.090393
13 238.10 2017-03-16 233.984467
14 237.03 2017-03-17 234.309448
15 236.32 2017-03-20 235.186676
16 233.58 2017-03-21 238.271561
17 233.05 2017-03-22 236.720673
18 233.60 2017-03-23 235.235657
19 232.96 2017-03-24 235.274506
20 231.61 2017-03-27 234.975571
21 233.14 2017-03-28 232.597351
22 234.72 2017-03-29 235.073471
23 235.27 2017-03-30 233.283554
24 235.68 2017-03-31 236.843430
25 233.91 2017-04-03 233.786209
26 234.56 2017-04-04 234.607452
27 234.54 2017-04-05 236.101517
28 234.42 2017-04-06 236.949188
29 234.64 2017-04-07 234.220001
30 234.73 2017-04-10 237.824661
31 233.34 2017-04-11 235.214157
32 233.77 2017-04-12 234.126328
33 232.51 2017-04-13 236.780426
34 232.51 2017-04-14 236.256302
35 232.88 2017-04-17 234.450073
36 233.08 2017-04-18 233.575821
37 233.18 2017-04-19 236.969101
38 234.13 2017-04-21 236.524261
39 234.56 2017-04-24 233.803986
40 237.81 2017-04-25 237.522141
41 238.35 2017-04-26 233.729202
42 237.98 2017-04-27 233.496613
43 237.93 2017-04-28 235.620361
44 238.20 2017-05-01 233.358170
45 238.30 2017-05-02 235.508835
46 237.70 2017-05-03 233.407562
47 237.78 2017-05-04 236.723862
48 238.68 2017-05-05 235.565948
49 239.17 2017-05-08 235.229599
--------------------------------------------------
Column=FOpen_5 accuracy=0.998177191116 mse=2.27659209481 num_predictions=55
COpen Date FOpen
0 236.64 2017-02-27 238.192352
1 236.67 2017-02-28 236.068298
2 238.39 2017-03-01 233.407837
3 239.56 2017-03-02 233.426804
4 238.17 2017-03-03 234.870193
5 237.50 2017-03-06 237.075485
6 237.71 2017-03-07 237.380112
7 237.34 2017-03-08 234.219711
8 236.70 2017-03-09 237.943634
9 237.97 2017-03-10 233.489441
10 237.62 2017-03-13 235.329956
11 237.18 2017-03-14 235.384659
12 237.56 2017-03-15 234.191605
13 239.11 2017-03-16 234.631668
14 237.75 2017-03-17 238.205200
15 237.03 2017-03-20 234.193771
16 237.47 2017-03-21 236.639816
17 233.77 2017-03-22 236.817902
18 234.00 2017-03-23 236.664276
19 234.38 2017-03-24 234.296265
20 231.93 2017-03-27 237.186676
21 233.27 2017-03-28 237.978928
22 234.99 2017-03-29 235.493591
23 235.47 2017-03-30 234.852448
24 235.90 2017-03-31 235.815125
25 235.80 2017-04-03 233.627853
26 234.95 2017-04-04 234.599976
27 236.26 2017-04-05 234.378769
28 234.94 2017-04-06 235.113602
29 235.15 2017-04-07 236.804703
30 235.36 2017-04-10 234.353027
31 234.90 2017-04-11 235.246429
32 234.74 2017-04-12 233.361099
33 233.64 2017-04-13 237.659561
34 233.64 2017-04-14 237.646362
35 233.11 2017-04-17 234.575958
36 233.72 2017-04-18 237.595978
37 234.52 2017-04-19 236.852570
38 235.25 2017-04-21 234.651871
39 237.18 2017-04-24 234.744324
40 237.91 2017-04-25 237.875153
41 238.51 2017-04-26 237.987625
42 238.77 2017-04-27 236.527328
43 238.90 2017-04-28 233.824768
44 238.68 2017-05-01 237.369019
45 238.84 2017-05-02 235.249466
46 238.29 2017-05-03 234.652802
47 238.83 2017-05-04 234.766861
48 239.19 2017-05-05 235.063766
49 239.75 2017-05-08 234.963684
50 239.96 2017-05-09 236.949936
51 239.96 2017-05-10 236.464096
52 239.35 2017-05-11 234.053070
53 239.09 2017-05-12 234.455276
54 239.47 2017-05-15 236.910614
--------------------------------------------------
Column=FLow_15 accuracy=0.99868083036 mse=1.15577788337 num_predictions=45
CLow Date FLow
0 236.35 2017-02-27 237.317108
1 236.02 2017-02-28 236.028763
2 238.37 2017-03-01 236.467682
3 238.21 2017-03-02 235.270294
4 237.73 2017-03-03 237.266052
5 237.01 2017-03-06 236.366791
6 237.71 2017-03-07 233.293579
7 236.40 2017-03-08 234.402542
8 235.74 2017-03-09 235.478241
9 236.59 2017-03-10 233.113327
10 237.24 2017-03-13 233.244949
11 236.19 2017-03-14 235.251572
12 237.29 2017-03-15 237.382629
13 238.10 2017-03-16 232.962952
14 237.03 2017-03-17 234.481262
15 236.32 2017-03-20 234.740311
16 233.58 2017-03-21 235.420532
17 233.05 2017-03-22 234.751709
18 233.60 2017-03-23 235.780365
19 232.96 2017-03-24 234.348801
20 231.61 2017-03-27 237.593353
21 233.14 2017-03-28 237.389633
22 234.72 2017-03-29 237.485062
23 235.27 2017-03-30 236.241272
24 235.68 2017-03-31 237.196823
25 233.91 2017-04-03 236.735779
26 234.56 2017-04-04 232.439789
27 234.54 2017-04-05 234.444763
28 234.42 2017-04-06 237.228287
29 234.64 2017-04-07 235.283813
30 234.73 2017-04-10 237.159470
31 233.34 2017-04-11 233.135452
32 233.77 2017-04-12 232.638535
33 232.51 2017-04-13 237.368668
34 232.51 2017-04-14 233.268280
35 232.88 2017-04-17 237.489105
36 233.08 2017-04-18 235.493378
37 233.18 2017-04-19 232.950668
38 234.13 2017-04-21 232.923477
39 234.56 2017-04-24 236.022293
40 237.81 2017-04-25 233.592133
41 238.35 2017-04-26 234.243805
42 237.98 2017-04-27 236.920731
43 237.93 2017-04-28 235.522476
44 238.20 2017-05-01 236.753662
--------------------------------------------------
Column=FHigh_10 accuracy=0.998665333528 mse=3.19837957813 num_predictions=50
CHigh Date FHigh
0 237.31 2017-02-27 238.116806
1 236.95 2017-02-28 238.073212
2 240.32 2017-03-01 238.702911
3 239.57 2017-03-02 235.622223
4 238.61 2017-03-03 235.265427
5 238.12 2017-03-06 239.444489
6 237.71 2017-03-07 238.095032
7 237.64 2017-03-08 239.200668
8 237.24 2017-03-09 234.270569
9 238.02 2017-03-10 239.583817
10 237.86 2017-03-13 236.506302
11 237.24 2017-03-14 234.556320
12 239.44 2017-03-15 233.477707
13 239.20 2017-03-16 235.687210
14 237.97 2017-03-17 237.614822
15 237.36 2017-03-20 237.824234
16 237.61 2017-03-21 238.901703
17 234.61 2017-03-22 237.868256
18 235.34 2017-03-23 236.751755
19 235.04 2017-03-24 238.544586
20 233.92 2017-03-27 234.341949
21 235.81 2017-03-28 234.234421
22 235.81 2017-03-29 237.246872
23 236.52 2017-03-30 235.180527
24 236.51 2017-03-31 237.984360
25 236.03 2017-04-03 235.013977
26 235.58 2017-04-04 233.606445
27 237.39 2017-04-05 237.210892
28 236.04 2017-04-06 238.062119
29 236.00 2017-04-07 235.649841
30 236.26 2017-04-10 239.241165
31 235.18 2017-04-11 237.498764
32 234.96 2017-04-12 236.820740
33 234.49 2017-04-13 239.183060
34 234.49 2017-04-14 236.651443
35 234.57 2017-04-17 236.410843
36 234.49 2017-04-18 235.937378
37 234.95 2017-04-19 239.047195
38 235.31 2017-04-21 238.909409
39 237.41 2017-04-24 232.706223
40 238.95 2017-04-25 238.227890
41 239.53 2017-04-26 234.073669
42 238.95 2017-04-27 235.486847
43 238.93 2017-04-28 237.671921
44 239.17 2017-05-01 232.673920
45 238.98 2017-05-02 236.700531
46 238.88 2017-05-03 233.697357
47 238.92 2017-05-04 238.041336
48 239.72 2017-05-05 236.363312
49 239.92 2017-05-08 236.201157
--------------------------------------------------
Column=FHigh_15 accuracy=0.998541531411 mse=1.55878995885 num_predictions=45
CHigh Date FHigh
0 237.31 2017-02-27 237.879074
1 236.95 2017-02-28 237.778702
2 240.32 2017-03-01 238.200607
3 239.57 2017-03-02 236.513977
4 238.61 2017-03-03 238.252991
5 238.12 2017-03-06 237.138977
6 237.71 2017-03-07 235.989029
7 237.64 2017-03-08 237.235672
8 237.24 2017-03-09 236.047653
9 238.02 2017-03-10 235.282043
10 237.86 2017-03-13 235.183029
11 237.24 2017-03-14 236.718567
12 239.44 2017-03-15 237.555557
13 239.20 2017-03-16 234.484604
14 237.97 2017-03-17 236.761780
15 237.36 2017-03-20 235.900726
16 237.61 2017-03-21 236.543427
17 234.61 2017-03-22 236.587799
18 235.34 2017-03-23 236.514786
19 235.04 2017-03-24 236.510010
20 233.92 2017-03-27 238.011017
21 235.81 2017-03-28 237.714203
22 235.81 2017-03-29 237.909973
23 236.52 2017-03-30 236.461090
24 236.51 2017-03-31 237.961746
25 236.03 2017-04-03 237.754578
26 235.58 2017-04-04 234.507462
27 237.39 2017-04-05 235.964752
28 236.04 2017-04-06 238.192520
29 236.00 2017-04-07 237.197021
30 236.26 2017-04-10 237.994797
31 235.18 2017-04-11 234.291840
32 234.96 2017-04-12 234.716782
33 234.49 2017-04-13 238.371140
34 234.49 2017-04-14 234.810455
35 234.57 2017-04-17 238.171463
36 234.49 2017-04-18 236.883545
37 234.95 2017-04-19 235.012711
38 235.31 2017-04-21 234.795029
39 237.41 2017-04-24 236.749542
40 238.95 2017-04-25 234.834839
41 239.53 2017-04-26 236.515854
42 238.95 2017-04-27 237.048889
43 238.93 2017-04-28 236.079636
44 239.17 2017-05-01 237.461151
--------------------------------------------------
Column=FOpen_25 accuracy=0.99848460541 mse=1.73816066913 num_predictions=35
COpen Date FOpen
0 236.64 2017-02-27 237.816574
1 236.67 2017-02-28 235.636246
2 238.39 2017-03-01 238.271469
3 239.56 2017-03-02 236.959274
4 238.17 2017-03-03 237.465118
5 237.50 2017-03-06 237.865158
6 237.71 2017-03-07 236.993179
7 237.34 2017-03-08 234.584763
8 236.70 2017-03-09 233.244110
9 237.97 2017-03-10 237.703644
10 237.62 2017-03-13 234.808472
11 237.18 2017-03-14 235.239822
12 237.56 2017-03-15 234.950027
13 239.11 2017-03-16 235.803452
14 237.75 2017-03-17 238.045044
15 237.03 2017-03-20 236.011124
16 237.47 2017-03-21 238.083694
17 233.77 2017-03-22 235.486710
18 234.00 2017-03-23 234.097260
19 234.38 2017-03-24 235.818817
20 231.93 2017-03-27 237.592194
21 233.27 2017-03-28 234.536087
22 234.99 2017-03-29 236.317245
23 235.47 2017-03-30 235.314636
24 235.90 2017-03-31 237.147217
25 235.80 2017-04-03 236.712540
26 234.95 2017-04-04 237.701721
27 236.26 2017-04-05 234.547760
28 235.15 2017-04-07 237.012497
29 235.36 2017-04-10 237.901123
30 234.90 2017-04-11 237.447983
31 234.74 2017-04-12 235.423920
32 233.64 2017-04-13 237.207001
33 233.64 2017-04-14 234.435822
34 233.11 2017-04-17 235.358307
--------------------------------------------------
Column=FHigh_30 accuracy=0.998540184532 mse=1.89523682882 num_predictions=30
CHigh Date FHigh
0 237.31 2017-02-27 238.654388
1 236.95 2017-02-28 237.913757
2 240.32 2017-03-01 237.615326
3 239.57 2017-03-02 237.942841
4 238.61 2017-03-03 235.067856
5 238.12 2017-03-06 236.428833
6 237.64 2017-03-08 237.182312
7 237.24 2017-03-09 237.329041
8 238.02 2017-03-10 237.844208
9 237.86 2017-03-13 235.093140
10 237.24 2017-03-14 234.985718
11 239.44 2017-03-15 238.100098
12 239.20 2017-03-16 234.762924
13 237.97 2017-03-17 237.206512
14 237.36 2017-03-20 237.845367
15 237.61 2017-03-21 237.746216
16 234.61 2017-03-22 235.123657
17 235.34 2017-03-23 233.172958
18 235.04 2017-03-24 238.084290
19 233.92 2017-03-27 234.328537
20 235.81 2017-03-28 238.564850
21 235.81 2017-03-29 238.536072
22 236.52 2017-03-30 237.857056
23 236.51 2017-03-31 236.691391
24 236.03 2017-04-03 238.618271
25 235.58 2017-04-04 235.602158
26 237.39 2017-04-05 236.506805
27 236.04 2017-04-06 237.739807
28 236.00 2017-04-07 237.714905
29 236.26 2017-04-10 238.364258
--------------------------------------------------
Column=FOpen_20 accuracy=0.99825314721 mse=1.28484845039 num_predictions=40
COpen Date FOpen
0 236.64 2017-02-27 237.917526
1 236.67 2017-02-28 232.084137
2 238.39 2017-03-01 234.533905
3 239.56 2017-03-02 238.029938
4 238.17 2017-03-03 235.261230
5 237.50 2017-03-06 236.595505
6 237.71 2017-03-07 235.027740
7 237.34 2017-03-08 235.152695
8 236.70 2017-03-09 237.827057
9 237.97 2017-03-10 235.309494
10 237.62 2017-03-13 237.443634
11 237.18 2017-03-14 232.899292
12 237.56 2017-03-15 233.487885
13 239.11 2017-03-16 234.831863
14 237.75 2017-03-17 233.841690
15 237.03 2017-03-20 238.318741
16 237.47 2017-03-21 232.302078
17 233.77 2017-03-22 237.313400
18 234.00 2017-03-23 238.370819
19 234.38 2017-03-24 231.074432
20 231.93 2017-03-27 238.103638
21 233.27 2017-03-28 235.634750
22 234.99 2017-03-29 235.533920
23 235.47 2017-03-30 234.074921
24 235.90 2017-03-31 238.692703
25 235.80 2017-04-03 237.288986
26 234.95 2017-04-04 235.055527
27 236.26 2017-04-05 237.628601
28 234.94 2017-04-06 236.558319
29 235.15 2017-04-07 234.751251
30 235.36 2017-04-10 237.804810
31 234.90 2017-04-11 238.187592
32 234.74 2017-04-12 237.725464
33 233.64 2017-04-13 235.753250
34 233.64 2017-04-14 238.004852
35 233.11 2017-04-17 235.053726
36 233.72 2017-04-18 236.913940
37 234.52 2017-04-19 235.347778
38 235.25 2017-04-21 234.452515
39 237.18 2017-04-24 238.080124
--------------------------------------------------
Column=FVolume_10 accuracy=0.907178016782 mse=1.1376357129e+14 num_predictions=50
CVolume Date FVolume
0 56515440.0 2017-02-27 49050020.0
1 96961938.0 2017-02-28 67179288.0
2 149158170.0 2017-03-01 57029576.0
3 70245978.0 2017-03-02 74554640.0
4 81974300.0 2017-03-03 59417156.0
5 55391533.0 2017-03-06 70559072.0
6 393822.0 2017-03-07 95975664.0
7 78168795.0 2017-03-08 75058032.0
8 90683918.0 2017-03-09 80309808.0
9 81991652.0 2017-03-10 117453024.0
10 57256824.0 2017-03-13 70703744.0
11 59880778.0 2017-03-14 98328504.0
12 96081750.0 2017-03-15 109490584.0
13 78343951.0 2017-03-16 165140944.0
14 89002111.0 2017-03-17 166284832.0
15 52536979.0 2017-03-20 84392992.0
16 131809275.0 2017-03-21 68577128.0
17 97569204.0 2017-03-22 65943088.0
18 100410277.0 2017-03-23 78339784.0
19 112504853.0 2017-03-24 109490584.0
20 87454452.0 2017-03-27 93902768.0
21 93483915.0 2017-03-28 106212224.0
22 61950354.0 2017-03-29 67807328.0
23 56737890.0 2017-03-30 95485984.0
24 73733094.0 2017-03-31 51628672.0
25 85546486.0 2017-04-03 72781400.0
26 56466195.0 2017-04-04 96616176.0
27 108800604.0 2017-04-05 65563532.0
28 69135757.0 2017-04-06 56224084.0
29 74412311.0 2017-04-07 84796976.0
30 67615302.0 2017-04-10 69178496.0
31 88045276.0 2017-04-11 55412748.0
32 81864436.0 2017-04-12 120624648.0
33 92880394.0 2017-04-13 60878416.0
34 92880394.0 2017-04-14 67484416.0
35 68405367.0 2017-04-17 86180384.0
36 83225821.0 2017-04-18 103908352.0
37 68699868.0 2017-04-19 68893144.0
38 110389847.0 2017-04-21 55616128.0
39 119209877.0 2017-04-24 111632960.0
40 76698265.0 2017-04-25 64758252.0
41 84702455.0 2017-04-26 93897600.0
42 57410326.0 2017-04-27 77759200.0
43 63532845.0 2017-04-28 69928560.0
44 66882521.0 2017-05-01 129059168.0
45 57375732.0 2017-05-02 72307144.0
46 73137731.0 2017-05-03 86471680.0
47 61462732.0 2017-05-04 65864824.0
48 62001269.0 2017-05-05 77673384.0
49 48385730.0 2017-05-08 95289768.0
--------------------------------------------------
Column=FVolume_30 accuracy=0.930475108578 mse=1.29142162391e+14 num_predictions=30
CVolume Date FVolume
0 56515440.0 2017-02-27 154272176.0
1 96961938.0 2017-02-28 66343696.0
2 149158170.0 2017-03-01 45580672.0
3 70245978.0 2017-03-02 76765984.0
4 81974300.0 2017-03-03 74599040.0
5 55391533.0 2017-03-06 87300176.0
6 78168795.0 2017-03-08 95490720.0
7 90683918.0 2017-03-09 54597272.0
8 81991652.0 2017-03-10 70504896.0
9 57256824.0 2017-03-13 61960604.0
10 59880778.0 2017-03-14 58309852.0
11 96081750.0 2017-03-15 62352844.0
12 78343951.0 2017-03-16 59600176.0
13 89002111.0 2017-03-17 57883096.0
14 52536979.0 2017-03-20 74338800.0
15 131809275.0 2017-03-21 52386428.0
16 97569204.0 2017-03-22 73343280.0
17 100410277.0 2017-03-23 91719456.0
18 112504853.0 2017-03-24 60750428.0
19 87454452.0 2017-03-27 84663232.0
20 93483915.0 2017-03-28 63925408.0
21 61950354.0 2017-03-29 45640840.0
22 56737890.0 2017-03-30 82197232.0
23 73733094.0 2017-03-31 104147008.0
24 85546486.0 2017-04-03 57139560.0
25 56466195.0 2017-04-04 63300152.0
26 108800604.0 2017-04-05 111594736.0
27 69135757.0 2017-04-06 60577204.0
28 74412311.0 2017-04-07 87336096.0
29 67615302.0 2017-04-10 65963908.0
--------------------------------------------------
Column=FVolume_15 accuracy=0.881865035722 mse=1.50714346281e+14 num_predictions=45
CVolume Date FVolume
0 56515440.0 2017-02-27 58376660.0
1 96961938.0 2017-02-28 114653016.0
2 149158170.0 2017-03-01 68996536.0
3 70245978.0 2017-03-02 98230624.0
4 81974300.0 2017-03-03 53393920.0
5 55391533.0 2017-03-06 49446616.0
6 393822.0 2017-03-07 81472504.0
7 78168795.0 2017-03-08 100004472.0
8 90683918.0 2017-03-09 57248996.0
9 81991652.0 2017-03-10 71268888.0
10 57256824.0 2017-03-13 130026288.0
11 59880778.0 2017-03-14 99685816.0
12 96081750.0 2017-03-15 70750376.0
13 78343951.0 2017-03-16 94259024.0
14 89002111.0 2017-03-17 88722328.0
15 52536979.0 2017-03-20 67925600.0
16 131809275.0 2017-03-21 95118112.0
17 97569204.0 2017-03-22 56650368.0
18 100410277.0 2017-03-23 77995000.0
19 112504853.0 2017-03-24 90608664.0
20 87454452.0 2017-03-27 51847928.0
21 93483915.0 2017-03-28 64941312.0
22 61950354.0 2017-03-29 81022208.0
23 56737890.0 2017-03-30 44544140.0
24 73733094.0 2017-03-31 61810496.0
25 85546486.0 2017-04-03 64523312.0
26 56466195.0 2017-04-04 61715992.0
27 108800604.0 2017-04-05 51757008.0
28 69135757.0 2017-04-06 61607228.0
29 74412311.0 2017-04-07 128550936.0
30 67615302.0 2017-04-10 59937572.0
31 88045276.0 2017-04-11 73310760.0
32 81864436.0 2017-04-12 99653536.0
33 92880394.0 2017-04-13 61524240.0
34 92880394.0 2017-04-14 84920632.0
35 68405367.0 2017-04-17 60422616.0
36 83225821.0 2017-04-18 66732408.0
37 68699868.0 2017-04-19 53581332.0
38 110389847.0 2017-04-21 90439880.0
39 119209877.0 2017-04-24 103021816.0
40 76698265.0 2017-04-25 72200232.0
41 84702455.0 2017-04-26 60038452.0
42 57410326.0 2017-04-27 152863840.0
43 63532845.0 2017-04-28 68542696.0
44 66882521.0 2017-05-01 51390880.0
--------------------------------------------------
Column=FLow_20 accuracy=0.998418492538 mse=1.78245389644 num_predictions=40
CLow Date FLow
0 236.35 2017-02-27 236.516693
1 236.02 2017-02-28 232.719360
2 238.37 2017-03-01 234.400543
3 238.21 2017-03-02 236.472946
4 237.73 2017-03-03 235.309799
5 237.01 2017-03-06 235.155243
6 237.71 2017-03-07 234.749527
7 236.40 2017-03-08 234.886383
8 235.74 2017-03-09 237.387650
9 236.59 2017-03-10 234.101929
10 237.24 2017-03-13 236.617798
11 236.19 2017-03-14 234.345016
12 237.29 2017-03-15 233.537094
13 238.10 2017-03-16 234.894653
14 237.03 2017-03-17 233.620422
15 236.32 2017-03-20 236.831604
16 233.58 2017-03-21 233.621658
17 233.05 2017-03-22 237.020462
18 233.60 2017-03-23 237.310089
19 232.96 2017-03-24 230.665543
20 231.61 2017-03-27 236.240112
21 233.14 2017-03-28 235.566818
22 234.72 2017-03-29 234.099472
23 235.27 2017-03-30 233.678497
24 235.68 2017-03-31 236.371811
25 233.91 2017-04-03 235.914780
26 234.56 2017-04-04 234.261368
27 234.54 2017-04-05 236.242126
28 234.42 2017-04-06 236.769028
29 234.64 2017-04-07 235.570099
30 234.73 2017-04-10 237.313629
31 233.34 2017-04-11 236.876953
32 233.77 2017-04-12 237.146347
33 232.51 2017-04-13 235.157242
34 232.51 2017-04-14 236.425552
35 232.88 2017-04-17 233.392120
36 233.08 2017-04-18 235.693695
37 233.18 2017-04-19 235.397476
38 234.13 2017-04-21 232.840240
39 234.56 2017-04-24 236.576019
--------------------------------------------------
Column=FLow_25 accuracy=0.998601484905 mse=1.80604115137 num_predictions=35
CLow Date FLow
0 236.35 2017-02-27 236.538483
1 236.02 2017-02-28 233.490814
2 238.37 2017-03-01 236.478592
3 238.21 2017-03-02 237.355896
4 237.73 2017-03-03 236.198212
5 237.01 2017-03-06 237.066391
6 237.71 2017-03-07 236.528244
7 236.40 2017-03-08 233.600281
8 235.74 2017-03-09 232.905045
9 236.59 2017-03-10 236.264847
10 237.24 2017-03-13 234.175720
11 236.19 2017-03-14 235.140533
12 237.29 2017-03-15 234.296143
13 238.10 2017-03-16 234.653091
14 237.03 2017-03-17 236.870285
15 236.32 2017-03-20 236.012268
16 233.58 2017-03-21 235.739243
17 233.05 2017-03-22 234.380859
18 233.60 2017-03-23 233.576416
19 232.96 2017-03-24 234.317596
20 231.61 2017-03-27 236.920029
21 233.14 2017-03-28 236.372406
22 234.72 2017-03-29 235.406082
23 235.27 2017-03-30 235.085190
24 235.68 2017-03-31 236.167923
25 233.91 2017-04-03 237.649734
26 234.56 2017-04-04 236.838287
27 234.54 2017-04-05 233.396332
28 234.64 2017-04-07 236.977158
29 234.73 2017-04-10 236.686829
30 233.34 2017-04-11 236.656601
31 233.77 2017-04-12 234.104111
32 232.51 2017-04-13 237.367371
33 232.51 2017-04-14 233.513687
34 232.88 2017-04-17 234.464111
--------------------------------------------------
Column=FClose_30 accuracy=0.998366330784 mse=1.4605830873 num_predictions=30
CClose Date FClose
0 237.11 2017-02-27 236.038269
1 236.47 2017-02-28 238.432297
2 239.78 2017-03-01 238.986664
3 238.27 2017-03-02 237.920959
4 238.42 2017-03-03 234.579483
5 237.71 2017-03-06 236.220596
6 236.56 2017-03-08 235.853409
7 236.86 2017-03-09 238.039886
8 237.69 2017-03-10 237.464264
9 237.81 2017-03-13 233.276535
10 236.90 2017-03-14 233.017899
11 238.95 2017-03-15 237.794891
12 238.48 2017-03-16 233.586624
13 237.03 2017-03-17 236.831421
14 236.77 2017-03-20 237.214905
15 233.73 2017-03-21 238.634979
16 234.28 2017-03-22 234.244125
17 234.03 2017-03-23 233.014969
18 233.86 2017-03-24 238.156143
19 233.62 2017-03-27 231.728714
20 235.32 2017-03-28 237.859406
21 235.54 2017-03-29 238.588425
22 236.29 2017-03-30 237.473816
23 235.74 2017-03-31 237.031601
24 235.33 2017-04-03 237.774536
25 235.48 2017-04-04 233.693634
26 234.78 2017-04-05 235.902466
27 235.44 2017-04-06 236.567245
28 235.20 2017-04-07 236.367233
29 235.34 2017-04-10 238.269165
--------------------------------------------------
Column=FClose_15 accuracy=0.998329081518 mse=1.55153303883 num_predictions=45
CClose Date FClose
0 237.11 2017-02-27 237.514938
1 236.47 2017-02-28 236.349167
2 239.78 2017-03-01 237.232834
3 238.27 2017-03-02 235.958359
4 238.42 2017-03-03 237.573029
5 237.71 2017-03-06 235.776993
6 237.71 2017-03-07 233.257751
7 236.56 2017-03-08 234.547958
8 236.86 2017-03-09 236.204132
9 237.69 2017-03-10 234.440323
10 237.81 2017-03-13 235.245529
11 236.90 2017-03-14 235.861816
12 238.95 2017-03-15 237.665039
13 238.48 2017-03-16 234.893433
14 237.03 2017-03-17 235.102051
15 236.77 2017-03-20 234.524231
16 233.73 2017-03-21 235.744446
17 234.28 2017-03-22 236.093353
18 234.03 2017-03-23 235.905411
19 233.86 2017-03-24 234.976913
20 233.62 2017-03-27 238.391312
21 235.32 2017-03-28 238.297791
22 235.54 2017-03-29 237.847977
23 236.29 2017-03-30 235.961761
24 235.74 2017-03-31 237.490707
25 235.33 2017-04-03 236.914932
26 235.48 2017-04-04 234.191833
27 234.78 2017-04-05 235.793625
28 235.44 2017-04-06 238.008530
29 235.20 2017-04-07 235.931290
30 235.34 2017-04-10 237.555237
31 235.06 2017-04-11 234.810471
32 234.03 2017-04-12 233.186142
33 232.51 2017-04-13 238.099289
34 234.03 2017-04-14 234.776199
35 234.57 2017-04-17 238.332001
36 233.87 2017-04-18 235.706833
37 233.44 2017-04-19 235.013458
38 234.59 2017-04-21 234.846542
39 237.17 2017-04-24 235.933777
40 238.55 2017-04-25 234.720352
41 238.40 2017-04-26 235.560608
42 238.60 2017-04-27 235.922440
43 238.08 2017-04-28 235.809753
44 238.68 2017-05-01 237.397751
--------------------------------------------------
Column=FClose_10 accuracy=0.998336184677 mse=3.67349565905 num_predictions=50
CClose Date FClose
0 237.11 2017-02-27 236.719559
1 236.47 2017-02-28 235.765656
2 239.78 2017-03-01 236.341888
3 238.27 2017-03-02 234.455933
4 238.42 2017-03-03 234.006851
5 237.71 2017-03-06 239.427414
6 237.71 2017-03-07 237.065964
7 236.56 2017-03-08 238.872650
8 236.86 2017-03-09 232.399292
9 237.69 2017-03-10 238.442032
10 237.81 2017-03-13 235.934296
11 236.90 2017-03-14 234.442276
12 238.95 2017-03-15 232.743958
13 238.48 2017-03-16 232.663513
14 237.03 2017-03-17 236.680634
15 236.77 2017-03-20 236.893036
16 233.73 2017-03-21 238.207443
17 234.28 2017-03-22 238.872650
18 234.03 2017-03-23 235.746964
19 233.86 2017-03-24 235.897629
20 233.62 2017-03-27 233.996124
21 235.32 2017-03-28 233.275299
22 235.54 2017-03-29 236.156982
23 236.29 2017-03-30 235.352722
24 235.74 2017-03-31 236.800934
25 235.33 2017-04-03 234.667816
26 235.48 2017-04-04 234.142960
27 234.78 2017-04-05 236.400406
28 235.44 2017-04-06 236.916565
29 235.20 2017-04-07 236.253784
30 235.34 2017-04-10 238.985504
31 235.06 2017-04-11 237.273529
32 234.03 2017-04-12 235.987198
33 232.51 2017-04-13 238.770630
34 234.03 2017-04-14 236.319412
35 234.57 2017-04-17 235.258270
36 233.87 2017-04-18 236.158569
37 233.44 2017-04-19 238.818512
38 234.59 2017-04-21 240.200760
39 237.17 2017-04-24 232.249298
40 238.55 2017-04-25 236.471756
41 238.40 2017-04-26 232.417969
42 238.60 2017-04-27 235.982269
43 238.08 2017-04-28 235.568817
44 238.68 2017-05-01 233.334167
45 238.77 2017-05-02 235.452698
46 238.48 2017-05-03 233.297287
47 238.76 2017-05-04 236.348892
48 239.70 2017-05-05 235.108383
49 239.66 2017-05-08 235.850815
--------------------------------------------------
Column=FHigh_5 accuracy=0.998197941367 mse=1.80702890975 num_predictions=55
CHigh Date FHigh
0 237.31 2017-02-27 237.312714
1 236.95 2017-02-28 236.647583
2 240.32 2017-03-01 235.408691
3 239.57 2017-03-02 235.412170
4 238.61 2017-03-03 236.592255
5 238.12 2017-03-06 236.942337
6 237.71 2017-03-07 237.805847
7 237.64 2017-03-08 235.711090
8 237.24 2017-03-09 238.066299
9 238.02 2017-03-10 234.807449
10 237.86 2017-03-13 236.510498
11 237.24 2017-03-14 236.937378
12 239.44 2017-03-15 235.610382
13 239.20 2017-03-16 235.125275
14 237.97 2017-03-17 237.789337
15 237.36 2017-03-20 235.500015
16 237.61 2017-03-21 236.543228
17 234.61 2017-03-22 238.143082
18 235.34 2017-03-23 237.400833
19 235.04 2017-03-24 234.776672
20 233.92 2017-03-27 238.981216
21 235.81 2017-03-28 237.997787
22 235.81 2017-03-29 237.007416
23 236.52 2017-03-30 234.874252
24 236.51 2017-03-31 236.905701
25 236.03 2017-04-03 236.064346
26 235.58 2017-04-04 235.724167
27 237.39 2017-04-05 235.085098
28 236.04 2017-04-06 237.424881
29 236.00 2017-04-07 237.308105
30 236.26 2017-04-10 236.749710
31 235.18 2017-04-11 235.989014
32 234.96 2017-04-12 235.574341
33 234.49 2017-04-13 237.225113
34 234.49 2017-04-14 236.837738
35 234.57 2017-04-17 234.830063
36 234.49 2017-04-18 237.252274
37 234.95 2017-04-19 237.460571
38 235.31 2017-04-21 236.181732
39 237.41 2017-04-24 238.032349
40 238.95 2017-04-25 237.591309
41 239.53 2017-04-26 238.456604
42 238.95 2017-04-27 238.121109
43 238.93 2017-04-28 235.849518
44 239.17 2017-05-01 238.451675
45 238.98 2017-05-02 237.717178
46 238.88 2017-05-03 237.120483
47 238.92 2017-05-04 236.130096
48 239.72 2017-05-05 236.190536
49 239.92 2017-05-08 235.578979
50 240.19 2017-05-09 236.762985
51 240.19 2017-05-10 237.850525
52 239.46 2017-05-11 235.434219
53 239.43 2017-05-12 235.138306
54 240.44 2017-05-15 236.792328
--------------------------------------------------
Column=FClose_5 accuracy=0.998150179646 mse=1.98280524804 num_predictions=55
CClose Date FClose
0 237.11 2017-02-27 236.434555
1 236.47 2017-02-28 236.510025
2 239.78 2017-03-01 234.041580
3 238.27 2017-03-02 232.594925
4 238.42 2017-03-03 235.686096
5 237.71 2017-03-06 236.494537
6 237.71 2017-03-07 235.777527
7 236.56 2017-03-08 234.728180
8 236.86 2017-03-09 237.905029
9 237.69 2017-03-10 231.179260
10 237.81 2017-03-13 236.145966
11 236.90 2017-03-14 235.847733
12 238.95 2017-03-15 234.502991
13 238.48 2017-03-16 234.118912
14 237.03 2017-03-17 237.207382
15 236.77 2017-03-20 234.724274
16 233.73 2017-03-21 236.441284
17 234.28 2017-03-22 237.680954
18 234.03 2017-03-23 236.381516
19 233.86 2017-03-24 235.307632
20 233.62 2017-03-27 237.970642
21 235.32 2017-03-28 237.230652
22 235.54 2017-03-29 236.871536
23 236.29 2017-03-30 232.475800
24 235.74 2017-03-31 237.286194
25 235.33 2017-04-03 235.197113
26 235.48 2017-04-04 235.476990
27 234.78 2017-04-05 233.553894
28 235.44 2017-04-06 236.822678
29 235.20 2017-04-07 238.217285
30 235.34 2017-04-10 236.220367
31 235.06 2017-04-11 234.675919
32 234.03 2017-04-12 233.628799
33 232.51 2017-04-13 237.408386
34 234.03 2017-04-14 237.107529
35 234.57 2017-04-17 233.886322
36 233.87 2017-04-18 237.400650
37 233.44 2017-04-19 237.057693
38 234.59 2017-04-21 234.789215
39 237.17 2017-04-24 234.993515
40 238.55 2017-04-25 236.682175
41 238.40 2017-04-26 237.013947
42 238.60 2017-04-27 236.689621
43 238.08 2017-04-28 235.828934
44 238.68 2017-05-01 236.930801
45 238.77 2017-05-02 236.634476
46 238.48 2017-05-03 235.985107
47 238.76 2017-05-04 235.990326
48 239.70 2017-05-05 236.880493
49 239.66 2017-05-08 234.437592
50 239.44 2017-05-09 237.104736
51 239.66 2017-05-10 237.244461
52 239.87 2017-05-11 233.662079
53 238.98 2017-05-12 233.010788
54 240.30 2017-05-15 236.858719

Get the Analysis Images


In [6]:
job_res = get_job_analysis(job_id, show_plots=True)


Getting analysis for job=558 url=https://redten.io/ml/analysis/558/
SUCCESS - GET Analysis Response Status=200 Reason=OK
Found Job=558 analysis
SPY-2-558 5-Days - Predictive Accuracy
Predicted Close 5 Days vs Actual Close 5 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12122_7c49dccd964548fc.png
---------------------------------------------------------------------------------------
SPY-2-558 5-Days - Predictive Accuracy
Predicted High 5 Days vs Actual High 5 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12121_5ae84fdafa034174.png
---------------------------------------------------------------------------------------
SPY-2-558 10-Days - Predictive Accuracy
Predicted Close 10 Days vs Actual Close 10 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12120_2d87f32772344921.png
---------------------------------------------------------------------------------------
SPY-2-558 15-Days - Predictive Accuracy
Predicted Close 15 Days vs Actual Close 15 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12119_c54a1003037d4eb3.png
---------------------------------------------------------------------------------------
SPY-2-558 30-Days - Predictive Accuracy
Predicted Close 30 Days vs Actual Close 30 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12118_192e7a3c4ba840fc.png
---------------------------------------------------------------------------------------
SPY-2-558 25-Days - Predictive Accuracy
Predicted Low 25 Days vs Actual Low 25 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12117_6b74cc14075f4f20.png
---------------------------------------------------------------------------------------
SPY-2-558 20-Days - Predictive Accuracy
Predicted Low 20 Days vs Actual Low 20 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12116_1fed112e3a4e42ff.png
---------------------------------------------------------------------------------------
SPY-2-558 15-Days - Predictive Accuracy
Predicted Volume 15 Days vs Actual Volume 15 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12115_795d9657c2534f76.png
---------------------------------------------------------------------------------------
SPY-2-558 30-Days - Predictive Accuracy
Predicted Volume 30 Days vs Actual Volume 30 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12114_c19203024ec84944.png
---------------------------------------------------------------------------------------
SPY-2-558 10-Days - Predictive Accuracy
Predicted Volume 10 Days vs Actual Volume 10 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12113_1b000088cf154451.png
---------------------------------------------------------------------------------------
SPY-2-558 20-Days - Predictive Accuracy
Predicted Open 20 Days vs Actual Open 20 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12112_c72e079979774d17.png
---------------------------------------------------------------------------------------
SPY-2-558 30-Days - Predictive Accuracy
Predicted High 30 Days vs Actual High 30 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12111_b0c21c4c4d0d4e7a.png
---------------------------------------------------------------------------------------
SPY-2-558 25-Days - Predictive Accuracy
Predicted Open 25 Days vs Actual Open 25 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12110_2a27883839464a6c.png
---------------------------------------------------------------------------------------
SPY-2-558 15-Days - Predictive Accuracy
Predicted High 15 Days vs Actual High 15 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12109_325e5fefa07a444d.png
---------------------------------------------------------------------------------------
SPY-2-558 10-Days - Predictive Accuracy
Predicted High 10 Days vs Actual High 10 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12108_7dfe5cad6dd341c2.png
---------------------------------------------------------------------------------------
SPY-2-558 15-Days - Predictive Accuracy
Predicted Low 15 Days vs Actual Low 15 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12107_2fcaa612a10346d9.png
---------------------------------------------------------------------------------------
SPY-2-558 5-Days - Predictive Accuracy
Predicted Open 5 Days vs Actual Open 5 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12106_29c5e186270348f4.png
---------------------------------------------------------------------------------------
SPY-2-558 10-Days - Predictive Accuracy
Predicted Low 10 Days vs Actual Low 10 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12105_7139a2dd4df5447b.png
---------------------------------------------------------------------------------------
SPY-2-558 25-Days - Predictive Accuracy
Predicted Close 25 Days vs Actual Close 25 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12104_74d4b72e2bd342d6.png
---------------------------------------------------------------------------------------
SPY-2-558 5-Days - Predictive Accuracy
Predicted Volume 5 Days vs Actual Volume 5 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12103_286431b3eabc4af5.png
---------------------------------------------------------------------------------------
SPY-2-558 20-Days - Predictive Accuracy
Predicted Close 20 Days vs Actual Close 20 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12102_b8f6174000f5455c.png
---------------------------------------------------------------------------------------
SPY-2-558 30-Days - Predictive Accuracy
Predicted Low 30 Days vs Actual Low 30 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12101_bfe9888ad15648f9.png
---------------------------------------------------------------------------------------
SPY-2-558 25-Days - Predictive Accuracy
Predicted Volume 25 Days vs Actual Volume 25 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12100_0761ea1b0f684ce9.png
---------------------------------------------------------------------------------------
SPY-2-558 20-Days - Predictive Accuracy
Predicted Volume 20 Days vs Actual Volume 20 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12099_8f73f41290794595.png
---------------------------------------------------------------------------------------
SPY-2-558 20-Days - Predictive Accuracy
Predicted High 20 Days vs Actual High 20 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12098_5fd3e28826754f24.png
---------------------------------------------------------------------------------------
SPY-2-558 30-Days - Predictive Accuracy
Predicted Open 30 Days vs Actual Open 30 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12097_214858c79e704f4c.png
---------------------------------------------------------------------------------------
SPY-2-558 25-Days - Predictive Accuracy
Predicted High 25 Days vs Actual High 25 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12096_168bcf41ff5142ad.png
---------------------------------------------------------------------------------------
SPY-2-558 15-Days - Predictive Accuracy
Predicted Open 15 Days vs Actual Open 15 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12095_de06fd5548504b41.png
---------------------------------------------------------------------------------------
SPY-2-558 5-Days - Predictive Accuracy
Predicted Low 5 Days vs Actual Low 5 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12094_354bb5267c694ba5.png
---------------------------------------------------------------------------------------
SPY-2-558 10-Days - Predictive Accuracy
Predicted Open 10 Days vs Actual Open 10 Days
URL: https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12093_e2ecdc03bc8a4f30.png

Get the Recent Machine Learning Jobs


In [7]:
user_token = user_login(rt_user, rt_pass, rt_url)
auth_headers = {
                "Authorization" : "JWT " + str(user_token)
            }
resource_url = rt_url + "/ml/run/"
query_params = {}
post_data = {}

# Get the ML Job
resource_url = rt_url + "/ml/jobs/"

lg("Running Get ML Job url=" + str(resource_url), 6)
get_response = requests.get(resource_url, params=query_params, data=post_data, headers=auth_headers)

if get_response.status_code != 201 and get_response.status_code != 200:
    lg("Failed with GET Response Status=" + str(get_response.status_code) + " Reason=" + str(get_response.reason), 0)
    lg("Details:\n" + str(get_response.text) + "\n", 0)
else:
    lg("SUCCESS - GET Response Status=" + str(get_response.status_code) + " Reason=" + str(get_response.reason)[0:10], 5)

    as_json = True
    record = {}
    if as_json:
        record = json.loads(get_response.text)
        lg(ppj(record))
# end of post for running an ML Job


Running Get ML Job url=https://redten.io/ml/jobs/
SUCCESS - GET Response Status=200 Reason=OK
{
    "jobs": [
        {
            "algo_name": "xgb-regressor",
            "control_state": "active",
            "created": "2017-05-26 08-02-13",
            "csv_file": "",
            "desc": "Forecast simulation - 2017-05-26 08:02:13",
            "ds_name": "SPY",
            "feature_column_names": [
                "FHigh",
                "FLow",
                "FOpen",
                "FClose",
                "FVolume"
            ],
            "id": 558,
            "ignore_features": [
                "Ticker",
                "Date",
                "FDate",
                "FPrice",
                "DcsnDate",
                "Decision"
            ],
            "images": [
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12122_7c49dccd964548fc.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12121_5ae84fdafa034174.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12120_2d87f32772344921.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12119_c54a1003037d4eb3.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12118_192e7a3c4ba840fc.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12117_6b74cc14075f4f20.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12116_1fed112e3a4e42ff.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12115_795d9657c2534f76.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12114_c19203024ec84944.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12113_1b000088cf154451.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12112_c72e079979774d17.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12111_b0c21c4c4d0d4e7a.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12110_2a27883839464a6c.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12109_325e5fefa07a444d.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12108_7dfe5cad6dd341c2.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12107_2fcaa612a10346d9.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12106_29c5e186270348f4.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12105_7139a2dd4df5447b.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12104_74d4b72e2bd342d6.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12103_286431b3eabc4af5.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12102_b8f6174000f5455c.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12101_bfe9888ad15648f9.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12100_0761ea1b0f684ce9.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12099_8f73f41290794595.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12098_5fd3e28826754f24.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12097_214858c79e704f4c.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12096_168bcf41ff5142ad.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12095_de06fd5548504b41.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12094_354bb5267c694ba5.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12093_e2ecdc03bc8a4f30.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170526/2_558_12092_954ad32f51ef47f4.png"
            ],
            "manifest_secret_key": "live_2_f3f31f711f1b4081862a1927be0758880c3b3f00e2848aeb2a7f68997b4e17",
            "manifest_sloc": "redten-models-west:SPY-2-558_manifest.json",
            "max_features": 10,
            "ml_type": "Playbook-UnitsAhead",
            "prediction_type": "forecast",
            "rloc": "",
            "sloc": "",
            "status": "completed",
            "target_column_name": "FClose",
            "target_column_values": [
                "GoodBuys",
                "BadBuys",
                "Not Finished"
            ],
            "title": "SPY Forecast v5 - 34d6bf175cad4aaf9234aa06e95f4e4",
            "units_ahead_set": [
                5,
                10,
                15,
                20,
                25,
                30
            ],
            "units_ahead_type": "Days",
            "updated": "2017-05-26 08-03-53",
            "version": 1
        },
        {
            "algo_name": "xgb-regressor",
            "control_state": "active",
            "created": "2017-05-23 22-48-19",
            "csv_file": "/opt/work/data/src/iris.csv",
            "desc": "This is a Description for - Hello World - 2017-05-23-22-48-19",
            "ds_name": "iris_regressor",
            "feature_column_names": [
                "SepalLength",
                "SepalWidth",
                "PetalLength",
                "PetalWidth",
                "ResultTargetValue"
            ],
            "id": 557,
            "ignore_features": [
                "ResultLabel"
            ],
            "images": [
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12091_4fdb0c2731074b5c.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12090_82e81ab1f9d54d09.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12089_db2ec909b1f84946.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12088_7316b1cc3a86413d.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12087_f475c7062acc4e03.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12086_b3ef117e966d45a4.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12085_87d25f7a41ba41d5.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12084_2b85954122e4452d.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12083_87091b7a48f64bf8.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12082_160492c43ac5415d.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12081_994b29e1e64b4c48.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12080_bacd0a3063e643b1.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12079_61d46d9599f6484d.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12078_8b49ca61b2984b6f.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12077_7c6bbe897e684ea3.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12076_ca83a370b4114a88.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12075_a0d64fc564fb4a3b.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12074_4591faca0ca54ecd.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12073_2a41ad8f67804a32.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12072_a4e071ac32fc4177.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12071_bc9e7fd426c444bf.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12070_7f5af55e914444aa.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12069_01ba1d04ca204f1a.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_557_12068_11ef9edc123542be.png"
            ],
            "manifest_secret_key": "live_2_9a5a34e96cca479e98ac8cc47d6fb53448557fa02d447c0bdcf167b8e4a645",
            "manifest_sloc": "redten-models-west:IRIS_REGRESSOR-2-557_manifest.json",
            "max_features": 10,
            "ml_type": "Predict with Filter",
            "prediction_type": "predict",
            "rloc": "",
            "sloc": "",
            "status": "completed",
            "target_column_name": "ResultLabel",
            "target_column_values": [
                "Iris-setosa",
                "Iris-versicolor",
                "Iris-virginica"
            ],
            "title": "Hello World - 2017-05-23-22-48-19",
            "units_ahead_set": [],
            "units_ahead_type": "",
            "updated": "2017-05-23 22-49-13",
            "version": 1
        },
        {
            "algo_name": "xgb-regressor",
            "control_state": "active",
            "created": "2017-05-23 22-24-09",
            "csv_file": "",
            "desc": "Forecast simulation - 2017-05-23 22:24:08",
            "ds_name": "SPY",
            "feature_column_names": [
                "FHigh",
                "FLow",
                "FOpen",
                "FClose",
                "FVolume"
            ],
            "id": 556,
            "ignore_features": [
                "Ticker",
                "Date",
                "FDate",
                "FPrice",
                "DcsnDate",
                "Decision"
            ],
            "images": [
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12067_a1d9483443844be5.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12066_bd242932692b48f1.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12065_c42e1ef3ecda4a50.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12064_4929471da7b848dc.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12063_3f0f9b7a91e6477d.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12062_50476dc42b95485a.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12061_087ef25764154835.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12060_c3c6ab2f04ae4669.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12059_974874fc3c0c4c4b.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12058_660025f112e04a51.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12057_ceced8e2215749e5.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12056_53ca82cd037c4ebb.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12055_76b743196398499f.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12054_72baf6fe1cff49dd.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12053_de8af6d2ab474eb4.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12052_26fbe6058c114619.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12051_184561aa8af3489a.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12050_430a5186b3d945e4.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12049_49d5e69d8c104eb2.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12048_7f90d455f7df4e2f.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12047_eba8de31f57e466a.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12046_8c585f82cd0948c5.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12045_db89689ffd49445c.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12044_f64ef76a00cb4bb9.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12043_ddeb69b9254b4595.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12042_152cb94cf47f4557.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12041_cf802c6530714ae7.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12040_a3e6aafa5e6241c4.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12039_dba3efb9b56c4bb5.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12038_c8fc6f8d4e294881.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_556_12037_0523b881d6a64baa.png"
            ],
            "manifest_secret_key": "live_2_ab288f9631d848b2a11cc4537728dc3d844c041bf934a6abdae5d6bbce3513",
            "manifest_sloc": "redten-models-west:SPY-2-556_manifest.json",
            "max_features": 10,
            "ml_type": "Playbook-UnitsAhead",
            "prediction_type": "forecast",
            "rloc": "",
            "sloc": "",
            "status": "completed",
            "target_column_name": "FClose",
            "target_column_values": [
                "GoodBuys",
                "BadBuys",
                "Not Finished"
            ],
            "title": "SPY Forecast v5 - 1394ae9d74c24d5886f9608d3dd6328",
            "units_ahead_set": [
                5,
                10,
                15,
                20,
                25,
                30
            ],
            "units_ahead_type": "Days",
            "updated": "2017-05-23 22-25-51",
            "version": 1
        },
        {
            "algo_name": "xgb-regressor",
            "control_state": "active",
            "created": "2017-05-23 22-00-11",
            "csv_file": "",
            "desc": "Forecast simulation - 2017-05-23 22:00:09",
            "ds_name": "SPY",
            "feature_column_names": [
                "FHigh",
                "FLow",
                "FOpen",
                "FClose",
                "FVolume"
            ],
            "id": 555,
            "ignore_features": [
                "Ticker",
                "Date",
                "FDate",
                "FPrice",
                "DcsnDate",
                "Decision"
            ],
            "images": [
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12036_b5cb1f3c8a554dd2.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12035_1cb38442136f4719.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12034_b74469132a794f8e.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12033_42cb601864994eb2.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12032_bf48ad03bdf7426e.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12031_19276b3843ea467e.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12030_bdcb901323804a6b.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12029_c7e56bc6a77748d6.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12028_f5d022f810ba4283.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12027_2620b246cfd54a68.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12026_bf5288b75d994cde.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12025_beec39f18bb84469.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12024_3a4a00f45ddb4472.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12023_916304a8e7184119.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12022_f335bee9141a4d39.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12021_3e6c529854ec47dd.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12020_5b859fd5b3ca417e.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12019_813dbe0a739d4aa4.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12018_dd20dc4f941946e4.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12017_3089d07bad8c4b85.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12016_029cb3266c46413d.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12015_5ee300ed25504c2b.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12014_5d8cfd12bc7f49bf.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12013_3b28fd99ff354edb.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12012_1a9331b9b8e24145.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12011_f65ee56325394208.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12010_17b819f13e914bfa.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12009_c70ec6e3690e466c.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12008_6e0222bdc5ab4adf.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12007_981709fca5e94062.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_555_12006_75950bb597ae44cf.png"
            ],
            "manifest_secret_key": "live_2_9cf95533d27e4668a88641797cf512d7f6e274a4c954f66a2f9f53d20bee20",
            "manifest_sloc": "redten-models-west:SPY-2-555_manifest.json",
            "max_features": 10,
            "ml_type": "Playbook-UnitsAhead",
            "prediction_type": "forecast",
            "rloc": "",
            "sloc": "",
            "status": "completed",
            "target_column_name": "FClose",
            "target_column_values": [
                "GoodBuys",
                "BadBuys",
                "Not Finished"
            ],
            "title": "SPY Forecast v5 - 6b8e2873aaa84e3a8eadf5f260151c7",
            "units_ahead_set": [
                5,
                10,
                15,
                20,
                25,
                30
            ],
            "units_ahead_type": "Days",
            "updated": "2017-05-23 22-02-14",
            "version": 1
        },
        {
            "algo_name": "xgb-regressor",
            "control_state": "active",
            "created": "2017-05-23 17-54-29",
            "csv_file": "",
            "desc": "Forecast simulation - 2017-05-23 17:54:29",
            "ds_name": "SPY",
            "feature_column_names": [
                "FHigh",
                "FLow",
                "FOpen",
                "FClose",
                "FVolume"
            ],
            "id": 554,
            "ignore_features": [
                "Ticker",
                "Date",
                "FDate",
                "FPrice",
                "DcsnDate",
                "Decision"
            ],
            "images": [
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_12005_7c0f75253702492d.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_12004_85a9a625b71a46fa.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_12003_cb0f1d4d1e964042.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_12002_52aef96a1cc14982.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_12001_6124922b9b654a5f.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_12000_5b543ec29d73467d.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11999_5b4cd833dfce4890.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11998_a6d2b4f739604933.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11997_c27870831e55409a.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11996_28dcadfdcf5a4883.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11995_bfc6066c46714d71.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11994_384c006cf37e43c1.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11993_c3359d2630f746c0.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11992_8524dc5a2d4e4185.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11991_ce5fafef07ac4d22.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11990_cd029f4e120c4aba.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11989_22ab6a277b374201.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11988_f8e0267d93ee477c.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11987_c3381ff13ec64906.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11986_dfa69223565445ff.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11985_da5aa1505dc44776.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11984_8980ae40b6d944b5.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11983_6c550ae6bc9c43ff.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11982_a3fbf30ee01144f6.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11981_c5c10073f6e04e0d.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11980_606d78feecf84b27.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11979_5b6dc14929654885.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11978_f2e3475eb2ff48d4.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11977_ff15925c6cd3433c.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11976_2117953502474673.png",
                "https://rt-media.s3.amazonaws.com/media/imagesml/20170523/2_554_11975_723d049fdaf64c5b.png"
            ],
            "manifest_secret_key": "live_2_c0906bd2b8dc438eaf3c06415ccbf5705b4b2e10f744b9abcf036377d81b62",
            "manifest_sloc": "redten-models-west:SPY-2-554_manifest.json",
            "max_features": 10,
            "ml_type": "Playbook-UnitsAhead",
            "prediction_type": "forecast",
            "rloc": "",
            "sloc": "",
            "status": "completed",
            "target_column_name": "FClose",
            "target_column_values": [
                "GoodBuys",
                "BadBuys",
                "Not Finished"
            ],
            "title": "SPY Forecast v5 - 739024c2c4f745c89c978040405bf22",
            "units_ahead_set": [
                5,
                10,
                15,
                20,
                25,
                30
            ],
            "units_ahead_type": "Days",
            "updated": "2017-05-23 18-01-41",
            "version": 1
        }
    ]
}

Redis Machine Learning Manifest

Jobs use a manifest to prevent concurrent jobs in-flight and models from colliding between users and historical machine learning jobs

A manifest contains:

  1. A dictionary of Redis model locations
  2. S3 archival locations
  3. Tracking data for import and export across environments
  4. Decoupled large model files (8gb files in S3) from the tracking and deployment

In [8]:
job_manifest = get_job_cache_manifest(job_report)
lg(ppj(job_manifest))


{
    "11811": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_0",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11811_modelkey_MD_SPY-2-558_cd1d9a_0.cache.pickle.zlib",
        "target": "FOpen_10",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11812": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_1",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11812_modelkey_MD_SPY-2-558_cd1d9a_1.cache.pickle.zlib",
        "target": "FLow_5",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11813": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_2",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11813_modelkey_MD_SPY-2-558_cd1d9a_2.cache.pickle.zlib",
        "target": "FOpen_15",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11814": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_3",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11814_modelkey_MD_SPY-2-558_cd1d9a_3.cache.pickle.zlib",
        "target": "FHigh_25",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11815": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_4",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11815_modelkey_MD_SPY-2-558_cd1d9a_4.cache.pickle.zlib",
        "target": "FOpen_30",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11816": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_5",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11816_modelkey_MD_SPY-2-558_cd1d9a_5.cache.pickle.zlib",
        "target": "FHigh_20",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11817": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_6",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11817_modelkey_MD_SPY-2-558_cd1d9a_6.cache.pickle.zlib",
        "target": "FVolume_20",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11818": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_7",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11818_modelkey_MD_SPY-2-558_cd1d9a_7.cache.pickle.zlib",
        "target": "FVolume_25",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11819": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_8",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11819_modelkey_MD_SPY-2-558_cd1d9a_8.cache.pickle.zlib",
        "target": "FLow_30",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11820": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_9",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11820_modelkey_MD_SPY-2-558_cd1d9a_9.cache.pickle.zlib",
        "target": "FClose_20",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11821": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_10",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11821_modelkey_MD_SPY-2-558_cd1d9a_10.cache.pickle.zlib",
        "target": "FVolume_5",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11822": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_11",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11822_modelkey_MD_SPY-2-558_cd1d9a_11.cache.pickle.zlib",
        "target": "FClose_25",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11823": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_12",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11823_modelkey_MD_SPY-2-558_cd1d9a_12.cache.pickle.zlib",
        "target": "FLow_10",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11824": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_13",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11824_modelkey_MD_SPY-2-558_cd1d9a_13.cache.pickle.zlib",
        "target": "FOpen_5",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11825": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_14",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11825_modelkey_MD_SPY-2-558_cd1d9a_14.cache.pickle.zlib",
        "target": "FLow_15",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11826": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_15",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11826_modelkey_MD_SPY-2-558_cd1d9a_15.cache.pickle.zlib",
        "target": "FHigh_10",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11827": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_16",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11827_modelkey_MD_SPY-2-558_cd1d9a_16.cache.pickle.zlib",
        "target": "FHigh_15",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11828": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_17",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11828_modelkey_MD_SPY-2-558_cd1d9a_17.cache.pickle.zlib",
        "target": "FOpen_25",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11829": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_18",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11829_modelkey_MD_SPY-2-558_cd1d9a_18.cache.pickle.zlib",
        "target": "FHigh_30",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11830": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_19",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11830_modelkey_MD_SPY-2-558_cd1d9a_19.cache.pickle.zlib",
        "target": "FOpen_20",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11831": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_20",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11831_modelkey_MD_SPY-2-558_cd1d9a_20.cache.pickle.zlib",
        "target": "FVolume_10",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11832": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_21",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11832_modelkey_MD_SPY-2-558_cd1d9a_21.cache.pickle.zlib",
        "target": "FVolume_30",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11833": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_22",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11833_modelkey_MD_SPY-2-558_cd1d9a_22.cache.pickle.zlib",
        "target": "FVolume_15",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11834": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_23",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11834_modelkey_MD_SPY-2-558_cd1d9a_23.cache.pickle.zlib",
        "target": "FLow_20",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11835": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_24",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11835_modelkey_MD_SPY-2-558_cd1d9a_24.cache.pickle.zlib",
        "target": "FLow_25",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11836": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_25",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11836_modelkey_MD_SPY-2-558_cd1d9a_25.cache.pickle.zlib",
        "target": "FClose_30",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11837": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_26",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11837_modelkey_MD_SPY-2-558_cd1d9a_26.cache.pickle.zlib",
        "target": "FClose_15",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11838": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_27",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11838_modelkey_MD_SPY-2-558_cd1d9a_27.cache.pickle.zlib",
        "target": "FClose_10",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11839": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_28",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11839_modelkey_MD_SPY-2-558_cd1d9a_28.cache.pickle.zlib",
        "target": "FHigh_5",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    },
    "11840": {
        "rloc": "MODELS:_MD_SPY-2-558_cd1d9a_29",
        "sloc": "redten-models-west:rt_models_userid_2_job_558_train_450_jobresults_413_modelid_11840_modelkey_MD_SPY-2-558_cd1d9a_29.cache.pickle.zlib",
        "target": "FClose_5",
        "tracking_id": "ML_SPY-2-558_95aa375d14684a8182fe20b566090a3",
        "tracking_name": "SPY-2-558",
        "tracking_type": "UseTargetColAndUnits"
    }
}

Multiple Models stored in Redis

Here's how models are stored in the Redis machine learning data store

Conclusion

Today's talk focused on:

  1. Using Redis as a machine learning data store for housing 1000s of pre-trained models
  2. Streamlining model pipelines to automate build + train + predict + export/import using a REST API
  3. How time intensive and expensive it is to continually rebuild machine learning models from scratch
  4. The importance of tracking model accuracy and performance over time
  5. How using a system like Red10 can enable an organization or team of data scientists to quickly test datasets and new ideas without stomping on each other's work
  6. How this approach can make predictions from any dataset...not just stocks
  7. This can make predictions with lots of different technologies