Create a .env and include the following filling in your account URL/Username/Password
SNOWFLAKE_ACCOUNT =
SNOWFLAKE_USER =
SNOWFLAKE_PASSWORD =
SNOWFLAKE_ROLE = sysadmin
SNOWFLAKE_WAREHOUSE = compute_wh
SNOWFLAKE_DATABASE = snowpark
SNOWFLAKE_SCHEMA = titanic
Run the load_data notebook which will perform the following tasks
- Load Titanic dataset from Seaborn, uppercase the column names and convert to csv
- Put the CSV file into a Snowflake Internal Stage
- Create a Snowpark DataFrame from the CSV in the stage
- Write the Snowpark DataFrame to Snowflake as a table
Run the snowml notebook which will perform the following tasks
- Create a Snowpark DataFrame from the Titanic table
- Check Null values
- drop columns with high count of nulls
- Convert Fare datatype
- Impute Categorical columns with nulls
- One Hot Encode Categrocial Values
- Split into Test & Train
- Train an XGBOOST Classifier Model
- Perform predictions on test
- Return Accuracy, Precision, and Recall