mohammadian7/freqtrade
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: avoid DataFrame fragmentation in get_predictions_to_append

Browse Source 
Replace column-by-column DataFrame assignment with dict-based
construction. The previous approach triggered pandas
PerformanceWarning about DataFrame fragmentation when many
prediction columns and their corresponding mean/std columns
were added one at a time.

Also add defensive key checks (label in self.data["labels_mean"])
to prevent KeyError when custom models produce prediction columns
that don't have corresponding entries in labels_mean/labels_std.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
Briarion
2026-03-16 08:46:57 +03:00
parent 59a81dc07a
commit a8c887a5b0
1 changed files with 14 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
freqtrade/freqai/data_kitchen.py
+14 -8
View File
@@ -428,18 +428,24 @@ class FreqaiDataKitchen:
Get backtest prediction from current backtest period
"""
append_df = DataFrame()
# Build dict first and construct DataFrame once to avoid
# column-by-column assignment which causes DataFrame fragmentation
# and PerformanceWarning on large prediction sets.
append_dict: dict[str, Any] = {}
for label in predictions.columns:
append_df[label] = predictions[label]
if append_df[label].dtype == object:
append_dict[label] = predictions[label]
if predictions[label].dtype == object:
continue
if "labels_mean" in self.data:
append_df[f"{label}_mean"] = self.data["labels_mean"][label]
if "labels_std" in self.data:
append_df[f"{label}_std"] = self.data["labels_std"][label]
if "labels_mean" in self.data and label in self.data["labels_mean"]:
append_dict[f"{label}_mean"] = self.data["labels_mean"][label]
if "labels_std" in self.data and label in self.data["labels_std"]:
append_dict[f"{label}_std"] = self.data["labels_std"][label]
for extra_col in self.data["extra_returns_per_train"]:
append_df[f"{extra_col}"] = self.data["extra_returns_per_train"][extra_col]
append_dict[f"{extra_col}"] = self.data["extra_returns_per_train"][extra_col]
append_df = DataFrame(append_dict)
append_df["do_predict"] = do_predict
if self.freqai_config["feature_parameters"].get("DI_threshold", 0) > 0: