Skip to content

Releases: radicalbit/radicalbit-ai-monitoring

v1.2.0

10 Dec 09:46
87488ee
Compare
Choose a tag to compare

1.2.0 (2024-12-10)

Features

Bug Fixes

v1.1.0

31 Oct 12:34
c2713ec
Compare
Choose a tag to compare

1.1.0 (2024-10-31)

Features

Bug Fixes

v1.0.1

20 Aug 09:37
955ec86
Compare
Choose a tag to compare

1.0.1 (2024-08-20)

Features

  • add modelUUID as optional value (#163) (672d6b4)
  • api: add ability to use environment variables to set spark annotations and resources (#165) (c0843b3)
  • ui: Grafana Faro usage improvement (#164) (022e94c)

v1.0.0

05 Aug 15:20
28f485a
Compare
Choose a tag to compare

1.0.0 (2024-08-05)

Features

  • add check (#145) (ba65d21)
  • add log loss property for classification models (#131) (d744235)
  • add regression line to residuals metric (#111) (c2f5437)
  • added CHI2 drift to float columns when cardinality < 15 (#121) (62f3852)
  • added field type in drift return, removes threshold of 15 for numerical vatriables (#141) (b08e97d)
  • added field type inside drift data (#142) (a19cd20)
  • added new log loss metric for binary - current and reference (#129) (52fe9a7)
  • added PSI drift method to int columns (#114) (07746f5)
  • added regression line calculation (#109) (c5511ed)
  • align demo data (#136) (039960b)
  • api: add field type as column definition property (#133) (92342b9)
  • api: add update model features API (#143) (80875c2)
  • api: bump libraries and fix log (#139) (1bd820d)
  • changed indexing method, removed job tests (#126) (a2aa7c5)
  • improve chi2 (#128) (ce76c86)
  • new field_type added, fixed all tests accordingly (#140) (dd6e85b)
  • new PSI value, chi2 new method fixed (#134) (d59f33c)
  • refactor jobs to test output (#120) (67e94e9)
  • sdk: added update model features (#146) (f439c17)
  • sdk: improved column definition class (#138) (d126e15)
  • ui: add grafana faro on UI (#149) (eb440ec)
  • ui: add logLoss to modelQuality binary (#135) (c9d407a)
  • ui: add PSI label (#137) (3c5e0a6)
  • ui: Edit fieldType in Overview -> variables section (#147) (3f228b2)
  • ui: improve regression section and add residuals charts (#118) (ffa9947)
  • ui: refactor drift section with field type (#148) (433f8bd)
  • ui: remove counterLabel and hide element in header (#123) (c3a744c)
  • ui: update label and add legend on chart (#127) (762e1ac)
  • ui: use step 3 to select targets,features and predictions (#130) (7eeedae)

Bug Fixes

  • add schema in postgres (#122) (e8e143c)
  • added psi to enum drift dto (#124) (e601c2d)
  • edit regression line structure in model quality dto (#117) (bf9d612)
  • finished refactoring for withcolumn prefix (#132) (48c1730)
  • return value for regression line casted to float (#115) (d5d862d)
  • return values for regression line changed from points to coefficients (#116) (e475478)
  • sdk: fix documentation typos (#125) (92ade35)
  • ui: fix regression resize chart (#119) (43fc389)

v0.9.0

16 Jul 11:54
52ac792
Compare
Choose a tag to compare

0.9.0 (2024-07-16)

Features

  • add current model quality regression (#90) (65f4f65)
  • add current model quality to multiclass metrics (#71) (9cd152f)
  • add data quality for regression (#78) (21d9139)
  • add data quality multiclass current (#56) (ccb6b79)
  • add data quality regression for reference data (#63) (92e327f)
  • add latest current and reference job status to modelout (#59) (f062c1d)
  • add multiclass dto for model quality of reference part (#46) (20e2d6b)
  • add reference regression model quality (#81) (2ee6365)
  • add regression current statistics (#72) (d873f05)
  • add residuals metrics to regression demo model, edit docker compose in order to block the start of api and ui before the up&running of k3s container first (#108) (8f204d9)
  • add residuals to regression dto (#106) (6a662aa)
  • add statistics in current dataset for multiclass (#53) (3881f97)
  • add support calculations to data quality of classification (#82) (ac26194)
  • added drift for regression (#95) (460f79e)
  • added job for regression model quality (#51) (9bf435a)
  • added model quality metrics for current dataset for regression (#87) (6a63d26)
  • added residual calculations, refactor tests (#104) (e344365)
  • added residuals to regression current, fix current job for mq - regression reference (#105) (2d4c024)
  • added support to binary and multiclass for predictions (#84) (bc61f93)
  • edit docker compose to add init-data container (#99) (fb6a08a)
  • refactoring model quality dtos (#92) (85d6a64)
  • reference multiclass model quality (#49) (e0e68d2)
  • spark: added data quality for current regression (#80) (7b5f6b5)
  • spark: added reference statistics for regression (#55) (cdff426)
  • spark: model quality and drift for current - multiclass classification (#74) (14d2cb9)
  • ui: add model quality for multiclass and other improvement (#85) (e75d1ea)
  • ui: add reference multiclass model quality (#69) (b505692)
  • ui: add support model quality multiclass (#93) (af3ec85)
  • ui: implement chart for multi classification (#68) (bfcd273)
  • ui: implement regression reference current (#94) (0035b60)
  • ui: regression scaffolng (#73) (26ef8f7)
  • update wizard (#54) (379ee62)

Bug Fixes

  • check that prediction_proba is not None before checking its type (#61) (6942317)
  • define single structure for drift metrics (#67) (aed740b)
  • edit model quality regression dto and data quality regression dto (#86) (c1dbcb2)
  • filter NaN and None before model quality metrics (#70) (02659e7)
  • return value for model quality regression to json (#91) (ed4ffd6)
  • set correlation id column as optional (#88) (e9d2b99)
  • set optional field of current multiclass model quality (#77) (17bed6b)
  • ui: add missing template for regression modelType (#102) (1ff9332)
  • ui: fix polling issue (#110) ([18fdd58](https:...
Read more

v0.8.2

28 Jun 10:16
69782a4
Compare
Choose a tag to compare

0.8.2 (2024-06-28)

Bug Fixes

  • dockerfile ui environment variables (#40) (15f8bbe)
  • manage types for all the fiels (#47) (64218fa)
  • ui: Improve polling and other fixies (#36) (951905f)

Chores

  • add latest reference and current uuids to modelOut dto (#32) (f8fdb5a)
  • add statistics for multiclassification + refactoring and improvements (#35) (18588ea)
  • allow to select the modelType and update the valid types prediction array (#43) (0c2ad1a)
  • api: get all reference/current datasets not paginated (#26) (6967044)
  • change type from secondary-light to primary (#42) (f2af815)
  • data quality for multiclass + fix class metrics (#41) (7cc870a)
  • fix binary (#48) (1a66a72)
  • get all models not paginated (#27) (06bd2ef)
  • refactoring of metrics part (#38) (6cb8bbe)
  • sdk: added ability to set S3 endpoint url if needed (#34) (c6a8f14)
  • sdk: get all reference and current datasets (#31) (5b1832d)
  • ui: add multi-classification folder structure and refactoring binary folder structure (#45) (7285847)

v0.8.1

25 Jun 09:24
d81d86b
Compare
Choose a tag to compare

0.8.1 (2024-06-25)

Bug Fixes

  • removed all .count() and replaced with attribute in the class definition (#16) (b86c5ec)
  • ui: fix import path for DataQualityMetrics and ModelQualityMetrics in import-detail-modal (#24) (0be20d5)
  • ui: show filename in import-detail-modal header (#25) (99629c5)

Chores

  • force dataType to TABULAR (#7) (c9c37ff)
  • sdk: align reference metrics business models with API (#11) (0db1dfe)
  • sdk: apply ruff format and lint like done in API project (#18) (41fd0cf)
  • sdk: get current dataset statistics (#13) (dec2d05)
  • sdk: get current model quality metrics (#21) (74fef65)
  • sdk: get data quality for current dataset (#19) (fce83aa)
  • sdk: implement get current dataset drift (#15) (7513b39)
  • sdk: manage all metrics based on model type (#22) (8d2710e)

v0.8.0

20 Jun 10:06
e046d02
Compare
Choose a tag to compare

0.8.0 (2024-06-20)

Features