Snowflake Data Science

I have enrolled in a private Snowflake Data Science Training. Let me list what I learned from it.

  • SQL worksheets
  • Snowpark in notebook
mkdocs build
mkdocs serve
mkdocs gh-deploy --force

SQL Worksheets

ML functions:

  • forecast
  • anomaly_detection
  • classification
  • top_insights

Add object name into session

show parameters like 'SEARCH_PATH';

set cur_search_path = (select "value" from table(result_scan(-1)));
set new_search_path = (select $cur_search_path || ', snowflake.ml'); -- append `snowflake.ml` into search_path

alter session set search_path = $new_search_path;

-- now below two statements are interchangeable 
show snowflake.ml.forecast;
show forecast;

Snowpark Notebook

Snowpark Configuration & Snowflak-Spark Configuration

:exclamation: Attributes are different

  • Snowpark config 
    {
  "account":"",
  "user":"",
  "authenticator":"externalbrowser",
  "role":"",
  "warehouse":"",
  "database":"",
  "schema":""
}
  • Snowflake Spark config 
    {
  "sfURL":"",
  "sfRole":"",
  "sfWarehouse":"",
  "sfDatabase":"",
  "sfSchema":"",
  "sfUser":"",
  "sfPassword":"",
  "authenticator":"externalbrowser",
}