Skip to main content

2.4.0 release notes

This release of AI-Link introduces new optimized operations for programmatic interaction with AtScale and expands available methods of authorization:

  • Expanded CRUD operation support for data models, perspectives, calculated columns and dimensions.
  • Enabled OAuth based connection to Snowflake

Please refer to our API documentation for the latest syntax to use with AI-Link. See below for updates associated with this release.

Data Warehouse Support Updates​

  • Added support for Postgres data warehouses.
  • Added clear_auth function to all warehouses, allowing for quick clearing of sensitive authentication information.
  • Added token parameter to Snowflake object constructor to support OAuth based connection

New Python Helper Functions for Programmatic Interaction​

  • CRUD operation support: added additional functions to interact with various objects in the semantic layer.
  • Creating DataModel: Project now has a function to create a blank DataModel, DataModel now has a function to clone itself.
  • Perspective updating: allows user to edit what is hidden by a given perspective.

Non-Functional Updates​

  • UX Quality of Life Improvements: additional customizations for the creation and management of aggs around udfs
  • Bug fixes: bug fixes addressing roleplaying and features with different key/value columns when joining objects to the semantic layer

Changelog for Syntax Updates​

enums.py​

REMOVED CLASS:

  • LevelType
    • functionality moved to the TimeSteps class

UPDATED CLASS:

  • TimeSteps

    • merged in functionality from former class LevelType to avoid duplication

  • TimeLevels

    • renamed class variable val to atscale_value

NEW CLASS:

  • Dimension
    • an enum to represent the metadata of a dimension object for dmv queries.
    • this does not have direct customer use cases but is publicly visible.

connection.py::Connection​

UPDATED FUNCTIONS:

  • _submit_request

    • builds in a retry in the event of an internal server error

  • get_connected_schemas

    • parameter database is now required

  • get_connected_tables

    • parameters database and schema are now required

bigquery.py::BigQuery​

UPDATED FUNCTIONS:

  • __init__
    • additional optional parameter warehouse_id added.
      • this is the AtScale warehouse id to associate with the connection when writing tables

NEW FUNCTIONS:

  • clear_auth
    • removes sensitive information from the connection object

databricks.py::Databricks​

UPDATED FUNCTIONS:

  • __init__
    • additional optional parameter warehouse_id added.
      • this is the AtScale warehouse id to associate with the connection when writing tables

NEW FUNCTIONS:

  • clear_auth
    • removes sensitive information from the connection object

iris.py::Iris​

UPDATED FUNCTIONS:

  • __init__
    • additional optional parameter warehouse_id added.
      • this is the AtScale warehouse id to associate with the connection when writing tables

NEW FUNCTIONS:

  • clear_auth
    • removes sensitive information from the connection object

mssql.py::MSSQL​

UPDATED FUNCTIONS:

  • __init__
    • additional optional parameter warehouse_id added.
      • this is the AtScale warehouse id to associate with the connection when writing tables

NEW FUNCTIONS:

  • clear_auth
    • removes sensitive information from the connection object

postgres.py::Postgres​

NEW FUNCTIONS:

  • added initial support for postgres data warehouses

redshift.py::Redshift​

UPDATED FUNCTIONS:

  • __init__
    • additional optional parameter warehouse_id added.
      • this is the AtScale warehouse id to associate with the connection when writing tables

NEW FUNCTIONS:

  • clear_auth
    • removes sensitive information from the connection object

snowflake.py::Snowflake​

UPDATED FUNCTIONS:

  • __init__

    • additional optional parameter token added.

      • adds in support for OAuth based authentication to Snowflake

    • additional optional parameter warehouse_id added.

      • this is the AtScale warehouse id to associate with the connection when writing tables

    • optional parameter private_key adjusted.

      • this parameter now exclusively accepts DER format

NEW FUNCTIONS:

  • token

    • getters and setters added for new parameter

  • clear_auth

    • removes sensitive information from the connection object

Synapse.py::synapse​

UPDATED FUNCTIONS:

  • __init__
    • additional optional parameter warehouse_id added.
      • this is the AtScale warehouse id to associate with the connection when writing tables

NEW FUNCTIONS:

  • clear_auth
    • removes sensitive information from the connection object

sql_connection.py::SQLConnection​

UPDATED FUNCTIONS:

  • __init__
    • additional optional parameter warehouse_id added.
      • this is the AtScale warehouse id to associate with the connection when writing tables

NEW FUNCTIONS:

  • clear_auth
    • removes sensitive information from the connection object

sqlalchemy_connection.py::SQLAlchemyConnection​

UPDATED FUNCTIONS:

  • __init__
    • additional optional parameter warehouse_id added.
      • this is the AtScale warehouse id to associate with the connection when writing tables

NEW FUNCTIONS:

  • clear_auth
    • removes sensitive information from the connection object

data_model.py::DataModel​

REMOVED FUNCTIONS:

  • get_column_names
    • removed in place of the more general get_columns function

UPDATED FUNCTIONS:

  • get_features

    • fixed a bug where the atscale_type field incorrectly returned Standard when use_unpublished set to True
    • secondary_attribute flag now included in response
    • add in checks to block running this function from a perspective with the use_published parameter set to False
    • fixed bug where passing use_published as False would return all calculated_measures in the project, instead of those in the data_model.

  • add_table

    • added parameter allow_aggregates which allows users to specify if aggregates will be built off of this dataset
    • updated parameters schema and database to now be required

  • fixed bug where variable referenced before declaration. This effects the following functions

    • writeback_spark
    • writeback_spark_to_spark

NEW FUNCTIONS:

  • clone

    • creates a clone of the current DataModel in the same project

  • get_columns

    • returns a dictionary of metadata for all the visible columns in a given dataset

  • get_dimensions

    • returns a dictionary of metadata for all dimensions in a DataModel

  • updated_calculated_columns

    • updates the sql expression for a calculated column

  • update_perspective

    • updates a perspective to hide the provided inputs

data_model_helpers.py​

UPDATED FUNCTIONS:

  • _check_joins

    • now prompts user when a join_column cannot be automatically mapped to a field or value column.

    • impacts the following functions

      • DataModel.add_query_dataset
      • DataModel.writeback
      • DataModel.writeback_spark
      • DataModel.writeback_spark_to_spark
      • DataModel.add_table
      • DataModel.create_dataset_relationship

model_utils.py​

UPDATED FUNCTIONS:

  • _create_dataset_relationship
    • fixed bug where writeback functions would fail with replacement as the dataset_id wasn't set correctly

db_utils.py​

UPDATED FUNCTIONS:

  • _get_key_cols
    • fixed bug where writeback functions could use wrong quotes when querying to join in missing key columns.

project.py::Project​

UPDATED FUNCTIONS:

  • get_snapshots
    • return type adjusted to be a list of dictionaries in an attempt to make response more useable

NEW FUNCTIONS:

  • create_data_model
    • creates a new empty DataModel in the current project

prediction_utils.py​

UPDATED FUNCTIONS:

  • join_udf

    • add parameters allow_aggregates and create_hinted_aggs so users can specify aggregate behavior

  • write_udf_to_qds

    • function renamed to write_snowpark_udf_to_qds as this function is only snowpark compatible
    • added publish parameter to the function