Removed incorrect parameter passed to pipeline classes in _add_baseline_pipelines
#941
Conversation
Codecov Report
@@ Coverage Diff @@
## main #941 +/- ##
==========================================
+ Coverage 99.86% 99.87% +0.01%
==========================================
Files 170 170
Lines 8645 8640 -5
==========================================
- Hits 8633 8629 -4
+ Misses 12 11 -1
Continue to review full report at Codecov.
|
evalml/automl/automl_search.py
Outdated
| strategy_dict = {"strategy": "mode"} | ||
| baseline = ModeBaselineBinaryPipeline(parameters={"Baseline Classifier": strategy_dict}) | ||
| elif self.problem_type == ProblemTypes.MULTICLASS: | ||
| strategy_dict = {"strategy": "random_weighted"} | ||
| strategy_dict = {"strategy": "mode"} | ||
| baseline = ModeBaselineMulticlassPipeline(parameters={"Baseline Classifier": strategy_dict}) | ||
| elif self.problem_type == ProblemTypes.REGRESSION: | ||
| strategy_dict = {"strategy": "mean"} |
There was a problem hiding this comment.
I decided to go with the classes we made and update the strategy dictionary. Alternative could be to create RandomWeightedBaselineBinaryPipeline etc 🤷
There was a problem hiding this comment.
Why not just update the ModeBaselineBinaryPipeline class definition with the right hyperparameters:
class ModeBaselineBinaryPipeline(BinaryClassificationPipeline):
"""Mode Baseline Pipeline for binary classification."""
custom_name = "Mode Baseline Binary Classification Pipeline"
component_graph = ["Baseline Classifier"]
custom_hyperparameters = {
"Baseline Classifier": {"strategy": ["mode"]}
}
There was a problem hiding this comment.
@freddyaboulton That's actually the definition of ModeBaselineBinaryPipeline right now! As to why we were using strategy_dict then... I'm going to have to refresh my memory 😂 . We could simply delete this and pass in no parameters, will circle back!
There was a problem hiding this comment.
@angela97lin This looks good to me. I have one question on the implementation but it's minor/non-blocking!
evalml/automl/automl_search.py
Outdated
| strategy_dict = {"strategy": "mode"} | ||
| baseline = ModeBaselineBinaryPipeline(parameters={"Baseline Classifier": strategy_dict}) | ||
| elif self.problem_type == ProblemTypes.MULTICLASS: | ||
| strategy_dict = {"strategy": "random_weighted"} | ||
| strategy_dict = {"strategy": "mode"} | ||
| baseline = ModeBaselineMulticlassPipeline(parameters={"Baseline Classifier": strategy_dict}) | ||
| elif self.problem_type == ProblemTypes.REGRESSION: | ||
| strategy_dict = {"strategy": "mean"} |
There was a problem hiding this comment.
Why not just update the ModeBaselineBinaryPipeline class definition with the right hyperparameters:
class ModeBaselineBinaryPipeline(BinaryClassificationPipeline):
"""Mode Baseline Pipeline for binary classification."""
custom_name = "Mode Baseline Binary Classification Pipeline"
component_graph = ["Baseline Classifier"]
custom_hyperparameters = {
"Baseline Classifier": {"strategy": ["mode"]}
}
angela97lin
changed the title
_add_baseline_pipelines_add_baseline_pipelines
| if num_pipelines == 0: | ||
| return True |
There was a problem hiding this comment.
Unrelated changes, but since we added in baseline pipelines, this will never be true and is not covered by codecov, so removing to pass codecov!
Closes #935