Skip to content

Latest commit

 

History

History
110 lines (56 loc) · 4.18 KB

README.md

File metadata and controls

110 lines (56 loc) · 4.18 KB

fer-gx

Validate data sources using Great Expectations Python Library.

Project context

This project is intended as a way to get to know "Great Expectations" Python library. It will rely on data coming from the NCY taxi data ( https://www.nyc.gov/site/tlc/about/tlc-trip-record-data.page ).

What can be achived?

By running the notebooks in this order:

  • a_suite_creation.ipynb

  • b_checkpoint_creation.ipynb

  • c_data_validation.ipynb

You will get the results of validating a dataset against different criteria. The pattern can be applied to build a testing suite against data assets in your pipeline, being them source data as well as transformed data.

Example:

Data source asset

Screenshot of NYC taxi data asset

What to expect from that asset

  • Expect column names in the asset to match this ordered list ["vendor_id", "pickup_datetime", "dropoff_datetime", "passenger_count", "trip_distance", "rate_code_id", "store_and_fwd_flag", "pickup_location_id", "dropoff_location_id", "payment_type", "fare_amount", "extra", "mta_tax", "tip_amount", "tolls_amount", "improvement_surcharge", "total_amount", "congestion_surcharge"]

  • Expect column rate_code_id values to be in the set {1}. This was intentionally made to generate unexpected values.

  • Expect 95% of the records to not be null for column vendor_id.

What we get from running validations

Overview of validation process

Overview of validation process

Column names match the ordered list

Column names match the ordered list

2.96% of rate_code_id column values do not conform to the values in set. Expectation was intentionally made to generate unexpected values

2.96% of rate_code_id column values do not conform to the values in set

The vendor_id column values are not null

vendor_id column values are not null

Notebooks and their content

I separated the scripts into three different files in order to assure things are getting saved to the context (Context and other terms are explained below). The scripts and their function are as follows:

a_suite_creation.ipynb

This notebook allows:

  1. The creation of a data context that is persistent across python sessions. "One of the primary responsibilities of the DataContext is managing CRUD operations for core GX objects" .(GX = Great Expectations).

  2. The creation of an expectation suite. An Expectation Suite is the placeholder for all the things you expect a certain data asset to conform to. Those things you expect are called Expectations.

    In this case, the expectation suite has the following expectations:

    • expect_table_columns_to_match_ordered_list

    • expect_column_values_to_be_in_set

    • expect_column_values_to_not_be_null

  3. The saving of the created expectation suite into the persistent context.

b_checkpoint_creation.ipynb

This notebook gets the context from its persistent storage and performs the following operations:

  1. Creates a Data Source and an asset related to it.

  2. Creates a Batch Request over that asset.

  3. Creates a Checkpoint that will perform validations from an Expectation Suite over the Batch Request. The Checkpoint is added to the context for being used when needed.

c_data_validation.ipynb

This notebook gets the context from its persistent storage and performs the following operations:

  1. Recovers the Checkpoint from the context.

  2. Gets a Result from running the Checkpoint.

  3. Shows the Result in HTML format.

Virtual environment

Python version: 3.11.4

Great expectations version: 0.17.12

The gxvenv virtual environment has been created to control the environment dependencies and added to .gitignore to prevent it from beeing copied to the repository. requirements.txt will contain the required modules and versions for this project.

To install dependencies for this project run:

pip install -r requirements.txt

What I learned

Great expectations library allows a "Continuous Integration" testing for ETL processes. It let you run tests on your source data as well as on transformed data. Then, any changes you make to your ETL process can be checked against your expectations by Great Expectations tests.