Changelog#

v1.0.7 (21/11/2023#

  • JOSS review changes

v1.0.6 (25/07/2023)#

  • Reduce stringency of dependencies

  • Update dependencies

  • Ensure datetime parsing can handle seconds

v1.0.5 (5/05/2023)#

  • Support Python 3.11 and add to CI testing

  • Update dependencies

  • Relicense nemseer under GPL3 and documentation under CC BY

v1.0.4 (5/12/2022)#

  • Version bump for Zenodo DOI

v1.0.3 (9/11/2022)#

  • Fix attrs dependency version

v1.0.2 (14/10/2022)#

Minor fixes:

  • Fix logging to use nemseer logger, not root logger

  • Suppress pandas Future Warning

v1.0.1 (27/09/2022)#

  • Where they exist, drop intervention periods when converting to xarray and provide user with warning

v1.0.0 (26/09/2022)#

  • Links to examples in all functions in Core Functionality of documentation

  • Detailed example of P5MIN demand forecast errors for 2021

v0.7.1 (20/09/2022)#

  • Various bug fixes from beta testing

    • Selective conversion of CSV to parquet

    • Correctly implement intervening date generation for filename generation

  • Example tidied up

  • Seconds can now be supplied in datetime inputs (as long as they are 00)

v0.7.0 (08/09/2022)#

  • Implement `processed_cache

  • Update documentation

v0.6.0 (06/09/2022)#

Major features#

  • Compile data to xarray Dataset

  • Smarter download_raw_data function

Minor additions#

  • Additional downcasting in conversion to parquet to reduce memory/file size

  • Update glossary with PD/STPASA run types

  • Additional info in quick start

Fixes#

  • Fix bug where data was mapped to enumerated table name

  • Remove duplicate rows when compiling

v0.5.0 (01/09/2022)#

  • Implement data compilation to pandas DataFrame

    • Add forecast-specific datetime validation to DataCompiler

    • Handle invalid/corrupted files during compilation

  • Add runtime generators

  • Expand testing for new functionality

  • Move constants to data module

  • Expand glossary and update README, API and quick start pages in docs

v0.4.0 (27/08/2022)#

  • Improve docs, including addition of glossary

  • Implement invalid/corrupt file handler for raw data downloading

v0.3.2 (25/08/2022)#

  • Expand testing, including testing all to check zip content exits for all possible URLs

  • Add hidden file handler for corrupted CSV/zips from NEMWeb

  • Improve doc referencing

  • Rename forecast_start/end to run_start/end

  • Add forecast-type datetime validation and corresponding testing

v0.3.1 (22/08/2022)#

  • Patch for functionality in v0.3.0

  • Refactor downloader to reduce redundant code

v0.3.0 (19/08/2022)#

  • Handle PREDISP_ALL_DATA (complete pre dispatch data for each month)

    • Associated testing

  • Update to docs

v0.2.1 (19/08/2022)#

  • Bug fix for Windows (PermissionError)

  • Switch date format to yyyy/mm/dd HH:MM for consistency with NEMOSIS and AEMO date format

  • Rename Loader to Query

  • Lazy parquet conversion

  • More testing, including doctesting of examples in Quick Start section of docs

v0.2.0 (16/08/2022)#

  • More sophisticated handling of raw data in raw_cache:

    • Check whether query data exists in raw_cache prior to downloading data

    • Progress bar for downloads

    • Cache data as parquet, with select columns parsed as datetime or category

  • Code base reorganised to facilitate function reuse

v0.1.0 (28/07/2022)#

  • First release of nemseer!

    • User feedback regarding available tables

    • Basic user input validation

    • Download and unzip files based on query