Exporting data from BigQuery to Google Cloud Storage

In this notebook, we export BigQuery data to GCS so that we can reuse our Keras model that was developed on CSV data.

Uncomment the following line if you are running the notebook locally:



In [ ]:

    
#%load_ext google.cloud.bigquery



In [ ]:

    
import os

from google.cloud import bigquery

Change the following cell as necessary:



In [ ]:

    
# Change with your own bucket and project below:
BUCKET =  "<BUCKET>"
PROJECT = "<PROJECT>"

OUTDIR = "gs://{bucket}/taxifare/data".format(bucket=BUCKET)

os.environ['BUCKET'] = BUCKET
os.environ['OUTDIR'] = OUTDIR
os.environ['PROJECT'] = PROJECT

Create BigQuery tables

If you haven not already created a BigQuery dataset for our data, run the following cell:



In [ ]:

    
bq = bigquery.Client(project = PROJECT)
dataset = bigquery.Dataset(bq.dataset("taxifare"))

try:
    bq.create_dataset(dataset)
    print("Dataset created")
except:
    print("Dataset already exists")

Let's create a table with 1 million examples.

Note that the order of columns is exactly what was in our CSV files.



In [ ]:

    
%%bigquery

CREATE OR REPLACE TABLE taxifare.feateng_training_data AS

SELECT
    (tolls_amount + fare_amount) AS fare_amount,
    pickup_datetime,
    pickup_longitude AS pickuplon,
    pickup_latitude AS pickuplat,
    dropoff_longitude AS dropofflon,
    dropoff_latitude AS dropofflat,
    passenger_count*1.0 AS passengers,
    'unused' AS key
FROM `nyc-tlc.yellow.trips`
WHERE ABS(MOD(FARM_FINGERPRINT(CAST(pickup_datetime AS STRING)), 1000)) = 1
AND
    trip_distance > 0
    AND fare_amount >= 2.5
    AND pickup_longitude > -78
    AND pickup_longitude < -70
    AND dropoff_longitude > -78
    AND dropoff_longitude < -70
    AND pickup_latitude > 37
    AND pickup_latitude < 45
    AND dropoff_latitude > 37
    AND dropoff_latitude < 45
    AND passenger_count > 0

Make the validation dataset be 1/10 the size of the training dataset.



In [ ]:

    
%%bigquery

CREATE OR REPLACE TABLE taxifare.feateng_valid_data AS

SELECT
    (tolls_amount + fare_amount) AS fare_amount,
    pickup_datetime,
    pickup_longitude AS pickuplon,
    pickup_latitude AS pickuplat,
    dropoff_longitude AS dropofflon,
    dropoff_latitude AS dropofflat,
    passenger_count*1.0 AS passengers,
    'unused' AS key
FROM `nyc-tlc.yellow.trips`
WHERE ABS(MOD(FARM_FINGERPRINT(CAST(pickup_datetime AS STRING)), 10000)) = 2
AND
    trip_distance > 0
    AND fare_amount >= 2.5
    AND pickup_longitude > -78
    AND pickup_longitude < -70
    AND dropoff_longitude > -78
    AND dropoff_longitude < -70
    AND pickup_latitude > 37
    AND pickup_latitude < 45
    AND dropoff_latitude > 37
    AND dropoff_latitude < 45
    AND passenger_count > 0

Export the tables as CSV files



In [ ]:

    
%%bash

echo "Deleting current contents of $OUTDIR"
gsutil -m -q rm -rf $OUTDIR

echo "Extracting training data to $OUTDIR"
bq --location=US extract \
   --destination_format CSV  \
   --field_delimiter "," --noprint_header \
   taxifare.feateng_training_data \
   $OUTDIR/taxi-train-*.csv

echo "Extracting validation data to $OUTDIR"
bq --location=US extract \
   --destination_format CSV  \
   --field_delimiter "," --noprint_header \
   taxifare.feateng_valid_data \
   $OUTDIR/taxi-valid-*.csv

gsutil ls -l $OUTDIR



In [ ]:

    
!gsutil cat gs://$BUCKET/taxifare/data/taxi-train-000000000000.csv | head -2

Copyright 2019 Google Inc. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.