In [1]:
# -*- coding: UTF-8 -*-
#%load_ext autoreload
%reload_ext autoreload
%autoreload 2
In [3]:
from __future__ import division
import tensorflow as tf
from os import path, remove
import numpy as np
import pandas as pd
import csv
from sklearn.model_selection import StratifiedShuffleSplit
from time import time
from matplotlib import pyplot as plt
import seaborn as sns
from mylibs.jupyter_notebook_helper import show_graph, renderStatsList, renderStatsCollection, \
renderStatsListWithLabels, renderStatsCollectionOfCrossValids, plot_res_gp, my_plot_convergence
from tensorflow.contrib import rnn
from tensorflow.contrib import learn
import shutil
from tensorflow.contrib.learn.python.learn import learn_runner
from mylibs.tf_helper import getDefaultGPUconfig
from sklearn.metrics import r2_score
from mylibs.py_helper import factors
from fastdtw import fastdtw
from collections import OrderedDict
from scipy.spatial.distance import euclidean
from statsmodels.tsa.stattools import coint
from common import get_or_run_nn
from data_providers.price_history_seq2seq_data_provider import PriceHistorySeq2SeqDataProvider
from data_providers.price_history_dataset_generator import PriceHistoryDatasetGenerator
from sklearn.metrics import mean_squared_error
from skopt.space.space import Integer, Real
from skopt import gp_minimize
from skopt.plots import plot_convergence
import pickle
import inspect
import dill
import sys
from models.model_21_price_history_seq2seq_dyn_dec_ins import PriceHistorySeq2SeqDynDecIns
from gp_opt.price_history_23_gp_opt import PriceHistory23GpOpt
from os.path import isdir
from cost_functions.huber_loss import huber_loss
In [4]:
dtype = tf.float32
seed = 16011984
random_state = np.random.RandomState(seed=seed)
config = getDefaultGPUconfig()
n_jobs = 1
%matplotlib inline
vocab_size is all the potential words you could have (classification for translation case) and max sequence length are the SAME thing
decoder RNN hidden units are usually same size as encoder RNN hidden units in translation but for our case it does not seem really to be a relationship there but we can experiment and find out later, not a priority thing right now
In [5]:
num_units = 400 #state size
input_len = 60
target_len = 30
batch_size = 64
with_EOS = False
In [6]:
total_size = 57994
train_size = 46400
test_size = 11584
In [7]:
data_folder = '../../../../Dropbox/data'
In [8]:
ph_data_path = '../data/price_history'
In [9]:
npz_full = ph_data_path + '/price_history_dp_60to30_57994.npz'
npz_train = ph_data_path + '/price_history_dp_60to30_57994_46400_train.npz'
npz_test = ph_data_path + '/price_history_dp_60to30_57994_11584_test.npz'
In [10]:
# dp = PriceHistorySeq2SeqDataProvider(npz_path=npz_train, batch_size=batch_size, with_EOS=with_EOS)
# dp.inputs.shape, dp.targets.shape
In [11]:
# aa, bb = dp.next()
# aa.shape, bb.shape
In [12]:
model = PriceHistorySeq2SeqDynDecIns(rng=random_state, dtype=dtype, config=config, with_EOS=with_EOS)
In [13]:
# graph = model.getGraph(batch_size=batch_size,
# num_units=num_units,
# input_len=input_len,
# target_len=target_len)
In [14]:
#show_graph(graph)
In [15]:
best_params = [500,
tf.nn.tanh,
0.0001,
0.62488034788862112,
0.001]
In [16]:
num_units, activation, lamda2, keep_prob_input, learning_rate = best_params
In [17]:
batch_size
Out[17]:
In [18]:
def experiment():
return model.run(npz_path=npz_train,
npz_test = npz_test,
epochs=100,
batch_size = batch_size,
num_units = num_units,
input_len=input_len,
target_len=target_len,
learning_rate = learning_rate,
preds_gather_enabled=True,
batch_norm_enabled = True,
activation = activation,
decoder_first_input = PriceHistorySeq2SeqDynDecIns.DECODER_FIRST_INPUT.ZEROS,
keep_prob_input = keep_prob_input,
lamda2 = lamda2,
)
In [19]:
#%%time
dyn_stats, preds_dict, targets = get_or_run_nn(experiment, filename='024_seq2seq_60to30_002',
nn_runs_folder= data_folder + '/nn_runs')
One epoch takes approximately 268 secs If we want to let it run for ~8 hours = 8 * 3600 / 268 ~= 107 epochs So let it run for 100 epochs and see how it behaves
In [20]:
dyn_stats.plotStats()
plt.show()
In [21]:
data_len = len(targets)
In [22]:
mses = np.empty(data_len)
for ii, (pred, target) in enumerate(zip(preds_dict.values(), targets.values())):
mses[ii] = mean_squared_error(pred, target)
In [23]:
np.mean(mses)
Out[23]:
In [24]:
huber_losses = np.empty(data_len)
for ii, (pred, target) in enumerate(zip(preds_dict.values(), targets.values())):
huber_losses[ii] = np.mean(huber_loss(pred, target))
In [25]:
np.mean(huber_losses)
Out[25]:
In [26]:
targets_arr = np.array(targets.values())
targets_arr.shape
Out[26]:
In [27]:
preds_arr = np.array(preds_dict.values())
preds_arr.shape
Out[27]:
In [28]:
np.mean(huber_loss(y_true=targets_arr, y_pred=preds_arr))
Out[28]:
In [29]:
r2_scores = [r2_score(y_true=targets[ind], y_pred=preds_dict[ind])
for ind in range(len(targets))]
In [30]:
ind = np.argmin(r2_scores)
ind
Out[30]:
In [31]:
reals = targets[ind]
preds = preds_dict[ind]
In [32]:
r2_score(y_true=reals, y_pred=preds)
Out[32]:
In [33]:
#sns.tsplot(data=dp.inputs[ind].flatten())
In [34]:
fig = plt.figure(figsize=(15,6))
plt.plot(reals, 'b')
plt.plot(preds, 'g')
plt.legend(['reals','preds'])
plt.show()
In [35]:
%%time
dtw_scores = [fastdtw(targets[ind], preds_dict[ind])[0]
for ind in range(len(targets))]
In [36]:
np.mean(dtw_scores)
Out[36]:
In [37]:
coint(preds, reals)
Out[37]:
In [52]:
cur_ind = np.random.randint(len(targets))
reals = targets[cur_ind]
preds = preds_dict[cur_ind]
fig = plt.figure(figsize=(15,6))
plt.plot(reals, 'b')
plt.plot(preds, 'g')
plt.legend(['reals','preds'])
plt.show()
In [ ]: