Adding the separate and surrogate regression method classes #379
Commits on Feb 15, 2024
Commits on Feb 16, 2024
-
-
-
-
-
-
-
put the original vignette into the long-vignette format as well
for speed-up + such that pkgdown don't need the torch stuff
-
-
-
-
-
Updated what to remove from shapr object when not saving vaeac model.…
… This is to reduce the .rds save files size and to remove time dependent random path names
-
-
-
-
-
-
Commits on Feb 19, 2024
-
Updated rbuildignore. Forgot to update after I split the vignettes in…
…to creatining fig/cache in separate sub folders
-
-
-
Commits on Feb 20, 2024
Commits on Feb 21, 2024
Commits on Feb 22, 2024
-
-
Adding logical indicating if we are to use regression and regression …
…cannot be combined with other approaches
-
-
Added
batch_prepare_vS_regression()which ensures that the grand co……alition is taken care off.
-
Added the regression_separate approach. Works without CV. Need to che…
…ck it more tommorow.
Commits on Feb 23, 2024
Commits on Feb 26, 2024
-
-
-
-
Fixed non-working approach check related to that regression approach …
…cannot be used in combined approaches.
Commits on Feb 27, 2024
Commits on Feb 28, 2024
Commits on Feb 29, 2024
-
-
-
-
-
Stopped using pipes %>% to make the code work in parallel using futur…
…e. It did not find the pipfunction before, and I did not want to load tidymodels in each parallel session.
-
-
-
Moved from right to left. I do not think I will ever be completely ha…
…ppy. But it is not important. Want left for " 1" and "10", but right for "0.1 " and "0.01"...
Commits on Mar 4, 2024
Commits on Mar 5, 2024
-
Started on potentially making tidymodels/workflow as a potential mode…
…l to be explained.
-
-
-
-
-
Commits on Mar 6, 2024
Commits on Mar 8, 2024
-
Updated regression train for surrogate regression to improve the cros…
…s validation by ensuring independence between training and evaluation data.
-
Changed to alwyas include the grand coalition when training surrogate…
… regression models.
-
-
-
Commits on Mar 11, 2024
-
Added option for
regression_tune_valuesto be a function. Needed fo……r, e.g., ranger and mtry, where mtry cannot be larger than n_features in each combination. The function can then take this into consideration and change the hyperparameter values for each coalition size
-
-
-
Commits on Mar 12, 2024
Commits on Mar 13, 2024
Commits on Mar 14, 2024
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Merge branch 'master' into Lars/shapr_regression # Please enter a commit message to explain why this merge is necessary, # especially if it merges an updated upstream into a topic branch. # # Lines starting with '#' will be ignored, and an empty message aborts # the commit.
-
-
-
-
-
-
-
-
-
-
-
Commits on Mar 15, 2024
Commits on Mar 18, 2024
-
Fixed such that we get the same surrogate model independent if the CV…
… was done in in parallel or sequential. Previously, we got the same optimal tuning values, but the random state was different, so for models with random fitting procedures, we got different models. Fixed by setting the seed right before fitting the model.
-
-
-
-
-
Fixed R-CMD to match how Martin did it
Merge remote-tracking branch 'refs/remotes/origin/Lars/shapr_regression' into Lars/shapr_regression # Conflicts: # .github/workflows/R-CMD-check.yaml
-
Specify that the regression approaches are only applicable when calle…
…d from R. One could technically use it from python to, but I am unsure how to let the user specify the tidymodel object.
-
Commits on Mar 20, 2024
Commits on Apr 2, 2024
-
I don't quite understand how it was possible to add them as they were in the local gitignore file...
Commits on Apr 3, 2024
Commits on Apr 4, 2024
-
-
-
-
-
-
-
Changed main vignette such that
explanationis not overwritten, and…… started adding workflows
Commits on Apr 8, 2024
Commits on Apr 9, 2024
Commits on Apr 10, 2024
-
-
Updated setup test snaps (changed text since we now use
regression.…… instead of underscore).
-
-
-
Updated such that the
regression.model,`regression.tune_values`, and `regression.recipe_func` parameter can be provided as strings
-
-
-
Commits on Apr 11, 2024
-
Merge branch 'master' into Lars/shapr_regression # Please enter a commit message to explain why this merge is necessary, # especially if it merges an updated upstream into a topic branch. # # Lines starting with '#' will be ignored, and an empty message aborts # the commit.
-
Commits on Apr 12, 2024
-
changed
batch_prepare_vStobatch_prepare_vS_MC_auxiliaryto make…… it clear that this function is a helper function to `batch_prepare_vS_MC`
-
-
Changed several things in python's explain:
0. Fixed such that we can use `regression_separate` and `regression_surrogate` in python. 1. Fixed indentation. 2. Specified where `.utils` is loaded from. 3. Added `**kwargs` to `explain()` such that we can actually send approach specific arguments. These must have the form `approach_parameter_name` in python, but added code to translate them to the R syntax `approach.parameter_name`. 4. Added documentation and some comments. 5. Refactored `compute_vS()` such that it splits between if we use MC-based or regression-based method. 6. As Python and pyr2 do not support tidymodels, we remove all objects related to `regression` in the `internal` list before converting it to python objects and returning them.
-
-
-
-
-
-
Commits on Apr 14, 2024
Commits on Apr 15, 2024
Commits on Apr 16, 2024
-
-
-
-
Martin: merged his updates into mine
Merge remote-tracking branch 'refs/remotes/origin/Lars/shapr_regression' into Lars/shapr_regression # Please enter a commit message to explain why this merge is necessary, # especially if it merges an updated upstream into a topic branch. # # Lines starting with '#' will be ignored, and an empty message aborts # the commit.
-
-
-
-
-
-
Made shaprpy-explain easier for Martin to look at.
timedeltais not…… used anywhere.
-