Python Idle
A junior data scientist proudly reports 99% accuracy on their housing price model. Your lead takes one look and says: "That's leakage." Data leakage happens when information from outside the training set bleeds into the model — scaling before splitting, features that encode the target, or temporal lookahead. Your job: find the bugs, fix the pipeline, and see what honest performance looks like.
~20 mindebug quest20640 rows
Loading Python runtime...
Goals: 3 tests
correct_score should be realistic (not inflated by leakage)
correct_score should be a realistic R2 value between 0.5 and 0.9
leakage_fixed flag should be set to True
Python loading...