Key Feature Additions and Changes | Teradata Package for Python 17.20 - Key Feature Additions and Changes - Teradata Vantage

Teradata® VantageCloud Lake

Deployment
VantageCloud
Edition
Lake
Product
Teradata Vantage
Published
January 2023
Language
English (United States)
Last Update
2024-02-17
dita:mapPath
phg1621910019905.ditamap
dita:ditavalPath
pny1626732985837.ditaval
dita:id
phg1621910019905

The following table lists the key feature additions and changes in the Teradata Package for Python, teradataml.

Date Release Description
Feburary 2024 17.20.00.07 Updated Open Analytics Framework APIs to support VantageCloud Lake use of Anaconda for building conda environments to run Python analytic workload on Open Analytics Framework:
  • Updated create_env() with new argument conda_env to specify whether the environment to be created is a conda environment or not.
  • Output of list environment APIs have a new column "conda" to show whether the environment is a conda environment or not.
  • Updated set_auth_token to address Open Analytics Login Issue with teradataml 17.20.00.05 and 17.20.00.06.
  • Updated list_user_envs() with new argument conda_env to specify whether to filter the conda environments when listing user environments.
January 2024 17.20.00.06
  • New teradataml DataFrame Column functions:
    • 19 new Bit Byte Manipulation Functions
    • 4 new Regular Expression Functions
    • 2 new Display Functions
  • New and updated Open Analytics Framework APIs:
    • Updated create_env() so user can create one or more user environments using the new argument template by providing specifications in template json file.
    • New UserEnv Class property models, and methods install_model() and uninstall_model() to list, install and uninstall models in user environment.
    • New UserEnv Class method snapshot() to take snapshot of user environment.
  • New BYOM function DataRobotPredict() to score the data in Vantage using the model trained externally in datarobot and stored in Vantage.
  • Updated DataFrame functions:
    • DataFrame.describe() method to accept argument statistics to specify the aggregate operation to perform.
    • DataFrame.sort() method to accept ColumnExpression, and enable sorting.
    • DataFrame.sample() method to support column stratification.
  • Updated general function view_log() to download the APPLY query logs.
  • Updated Analytics Database analytic functions so arguments which accept floating numbers will accept integers.
  • Updated DataFrame.plot() function to ignore the null values while plotting data.
Oct 2023 17.20.00.05
  • New hyperparameter tuning feature to determine the optimal set of hyperparameters for the given dataset and learning model.
    • GridSearch algorithm covers all possible parameter values to identify optimal hyperparameters.
    • RandomSearch algorithm performs random sampling on hyperparameter space to identify optimal hyperparameters.
  • New plotting feature to visualize analytic results.
  • New teradataml DataFrame functions:
    • DataFrame.plot() to generate plots on teradataml DataFrame.
    • DataFrame.itertuples() to iterate over teradataml DataFrame rows as namedtuples or list.
  • New teradataml GeoDataFrame function GeoDataFrame.plot() to generate plots on teradataml GeoDataFrame.
  • New BYOM function DataikuPredict() to score the data in Vantage using the model trained externally in Dataiku UI and stored in Vantage.
  • New teradataml DataFrame Column functions:
    • Regular Arithmetic Functions
    • Trigonometric Functions
    • Hyperbolic Functions
    • String Functions
  • New general function async_run_status() to check the status of asynchronous runs using unique run ids.
  • New teradataml configuration option configure.indb_install_location to specify the installation location of in-database Python package.
  • Updated Open Analytics Framework APIs:
    • set_auth_token() does not accept username and password anymore. Instead, function opens up a browser session and user should authenticate in browser.
    • User environments, files and libraries related APIs updated to support R environment.
  • Updated Unbounded Array Framework (UAF) function ArimaEstimate() to support for CSS algorithm via algorithm argument.
July 2023 17.20.00.04
  • teradataml is now compatible with SQLAlchemy 2.0.x.

    See SQLAlchemy Compatibility Considerations for important notes when using sqlalchemy 2.0 and later.

  • New utility function execute_sql() is added to execute SQL queries.

    See Running SQL Queries using teradataml and execute_sql() for details.

  • Native compatibility for Mac with ARM processors (without Rosetta).
  • Added support for floor division (//) between two teradataml DataFrame Columns.
  • New Analytics Database Analytic Functions:
    • GLMPerSegment()
    • GLMPredictPerSegment()
    • OneClassSVM()
    • OneClassSVMPredict()
    • SVM()
    • SVMPredict()
    • TargetEncodingFit()
    • TargetEncodingTransform()
    • TrainTestSplit()
    • WordEmbeddings()
    • XGBoost()
    • XGBoostPredict()
  • Added new Display option display.geometry_column_length to display the default length of geometry column in GeoDataFrame.
  • set_auth_token() function can generate the client id automatically based on org_id when not specified.
  • Updated Analytics Database Analytic Functions:
    • ColumnTransformer(): Does not allow list values for arguments onehotencoding_fit_data and ordinalencoding_fit_data
    • OrdinalEncodingFit(): Allows the list of values for arguments target_column, start_value, default_value; added arguments category_data, target_column_names, categories_column, ordinal_values_column.
    • OneHotEncodingFit(): Allows the list of values for arguments target_column, other_column; added arguments category_data, approach; target_columns, categories_column, category_counts.
May 2023 17.20.00.03
  • Removed the constraint that analytic functions must be imported after creating context.

    Analytic functions now can be imported regardless of context creation.

  • ReadNOS and WriteNOS now accept dictionary value for row_format and authorization arguments.
  • WriteNOS supports writing CSV files to external store.
  • Changes to the DataFrame.join:
    • Added new arguments lprefix and rprefix.

      Teradata recommends using these new arguments, instead of old arguments lsuffix and rsuffix. See join() Method for more details.

    • New and old affix arguments can now be used independently.
March 2023 17.20.00.02
  • New function set_auth_token() to set the JWT token automatically for using Open Analytics Framework APIs.
  • New display option display.suppress_vantage_runtime_warnings to suppress the VantageRuntimeWarning raised by teradataml, when set to True.
  • Updates to the following existing functions:
    • SimpleImputeFit function arguments stats_columns and stats are made to be optional.
    • ReadNOS function has new argument table_format, and argument full_scan is changed to scan_pct.
  • Added support of hash by and local order by to APPLY and DataFrame.apply().
January 2023 17.20.00.01
  • New Unbounded Array Framework (UAF) Time Series functions;
  • New input classes (TDSeries, TDGenSeries, TDMatrix, TDAnalyticResult) for UAF functions;
  • New DataFrame functions (pivot(), unpivot()) to rotate data to create easy-to-read DataFrames;
  • New DataFrame function drop_duplicate() to drop duplicate rows;
  • New DataFrame property is_art to check whether DataFrame is created on an Analytic Result Table (ART) or not.
  • Updates to the following existing functions:
    • Native Object Store (NOS) functions
    • display_analytic_functions() function
    • ColumnTransformer function
  • Analytics Database functions, UAF functions, NOS functions and BYOM functions are available only if underlying Vantage, which teradataml is connected to, supports the functions. And the functions should be imported only after the connection to Vantage is established.

    See Usage Notes when using Vantage Analytic Functions with teradataml, BYOM functions usage notes and NOS functions usage notes for more details.

November 2022 17.20.00.00
  • Support for use with VantageCloud Lake edition.
  • Support Analytics Database 17.20 Analytic Functions.
  • New Open Analytics feature APIs (user environment management functions, UserEnv Class functions, Apply Class functions) for use exclusively with Open Analytics Framework on VantageCloud Lake.
  • New DataFrame.apply method executes a user defined Python function on VantageCloud Lake.
  • New options (auth_token, base_url, certificate_file) for use with Open Analytics Framework on VantageCloud Lake.
  • New configuration option set_config_params set all config parameters in one go.
  • New BYOM function ONNXPredict() to score ONNX formatted models.
  • New database utility function list_td_reserved_keywords() to list Teradata reserved keyword.
  • New feature that special characters used in the password are encoded by default when using create_context.

    New optional argument url_encode can be set to 'False' to manually handle space and unreserved characters.