Skip to content

Fix: leaderboard fills nans #1432

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
eddiebergman merged 1 commit into from
Mar 26, 2022
+19 −0
Merged

Fix: leaderboard fills nans #1432

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 19 additions & 0 deletions autosklearn/estimators.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -937,6 +937,25 @@ def has_key(rv, key):
# tied together by ordering, might be better to store as tuple
for i, weight in enumerate(self.automl_.ensemble_.weights_):
(_, model_id, _) = self.automl_.ensemble_.identifiers_[i]

# We had issues where the model's in the ensembles are not in the runhistory
# collected. I have no clue why this is but to prevent failures, we fill
# the values with NaN
if model_id not in model_runs:
model_runs[model_id] = {
"model_id": model_id,
"seed": pd.NA,
"budget": pd.NA,
"duration": pd.NA,
"config_id": pd.NA,
"start_time": pd.NA,
"end_time": pd.NA,
"status": pd.NA,
"cost": pd.NA,
"train_loss": pd.NA,
"config_origin": pd.NA,
}

model_runs[model_id]["ensemble_weight"] = weight

# Filter out non-ensemble members if needed, else fill in a default
Expand Down