Releases: francescopisu/modelsight
Releases · francescopisu/modelsight
v0.4.0
24 Sep 16:51
Compare
Sorry, something went wrong.
No results found
build: install poetry then build
v0.3.6
24 Sep 13:35
Compare
Sorry, something went wrong.
No results found
v0.3.5
17 Sep 07:54
Compare
Sorry, something went wrong.
No results found
Documentation
Update README with graphical examples #10 (26102c4 )
v0.3.4
17 Sep 07:38
Compare
Sorry, something went wrong.
No results found
Documentation
Update example notebook + self-contained python script #3 (47d4cbc )
Update example notebook with average ROC curves plot of multiple binary classifiers (ef14075 )
v0.3.3
15 Sep 21:52
Compare
Sorry, something went wrong.
No results found
v0.3.2
15 Sep 17:50
Compare
Sorry, something went wrong.
No results found
v0.3.1
15 Sep 13:24
Compare
Sorry, something went wrong.
No results found
v0.3.0
14 Sep 11:07
Compare
Sorry, something went wrong.
No results found
Features
Added the possibility to make statistical comparison between areas under curve of different models in the ROC plot.
Updated the pytest's cv_results fixture to cross-validate multiple models.
v0.2.0
17 Jul 09:49
Compare
Sorry, something went wrong.
No results found
v0.2.0 (17/07/2023)
Feature
Introduced the curves module for producing average receiver-operating characteristic curved after cross-validation.
Added a custom dataclass type for the output of a cross-validation procedure: CVModellingOutput.
Tests
Added a fixture to simulate the output of a cross-validation procedure and added tests for the average receiver-operating characteristic curves.
Version 0.1.0
23 Jun 10:53
Compare
Sorry, something went wrong.
No results found
This is the first release of the modelsight package, featuring the Hosmer-Lemeshow plot to assess calibration of predicted probabilities.