Trifacta client that makes it easy to integrate Trifacta into your production and data science workflows

• Jupyter: Invoke Trifacta jobs from a Jupyter notebook and pass data back and forth between Jupyter and Trifacta
• Other Notebooks: Integrate Trifacta with Azure Databricks, Zepellin or any other notebook-style interface that supports Python
• Scripts: Automate Trifacta jobs and input/output using python scripts that can be easily executed from the command line or called from an external scheduler

This library makes it simple to do the following:

1. Connect to a Trifacta instance
2. Run a job
3. Download results to a csv file and view in pandas dataframe

Note that file uploads and downloads are performed using Amazon S3, using the boto3 API

#!pip install trifacta
import trifacta

If you need an access token, you can generate it as follows:

#Step 1: Connect to Trifacta by providing the URL and API Access Token
t = trifacta.Client('http://partnerdemo.amer.trifacta.net:3005', 'YOUR_ACCESS_TOKEN')

Make sure that you have run the job manually at least once

#Step 2: Run the job
t.run_job(23)

About to run job
{'sessionId': '9d339e65-8898-4165-871b-b9db848dc099', 'reason': 'JobStarted', 'jobGraph': {'vertices': [76, 77], 'edges': [{'source': 76, 'target': 77}]}, 'id': 42, 'jobs': {'data': [{'id': 76}, {'id': 77}]}}
2020-02-25 11:19:58.508231 InProgress
2020-02-25 11:20:03.700189 InProgress
2020-02-25 11:20:08.887794 Complete





True

%env AWS_PROFILE=trifacta_master_trial

env: AWS_PROFILE=trifacta_master_trial

#Step 3: Download results to a csv file and view in pandas dataframe
import boto3
s3 = boto3.client('s3', region_name='us-west-2')

s3.download_file(Bucket='trifacta-partnerdemo-trifactabucket-kkcpnw234feu',
                Key='trifacta/queryResults/admin@trifacta.local/MarketingAnalytics.csv',
                Filename='MarketingAnalytics.csv')

import pandas as pd
df = pd.read_csv('MarketingAnalytics.csv')

df.head()

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}