pandas/groupby-apply
| Rule | groupby-apply |
|---|---|
| Pack | pandas |
| Severity | warning |
| Fix | none, edit by hand |
| Library versions | every version |
groupby(...).apply with a lambda runs a Python function once for every group. Use agg or transform with named methods, idxmax for the row with the largest value, or sums of products for weighted means.
totals = df.groupby("region")["load"].apply(lambda group: group.sum())totals = df.groupby("region")["load"].sum()What the rule reports
Section titled “What the rule reports”The rule reports any groupby(...).apply with a lambda, which runs Python once per group. The rule leaves alone a lambda that sorts each group, or that builds a dict, a list or a joined text per group.
How to fix it
Section titled “How to fix it”xpd has no automatic fix, so make the change by hand.
The named reduction can change the name of the index, the dtype and the floating point rounding, so check the result before you replace apply.
Message
Section titled “Message”xpd prints this message with the issue. A name that starts with $ stands for code from your file.
groupby.apply runs a Python function once per group. Use agg or transform with named methods, df.loc[df.groupby(key)[col].idxmax()] for the row with the largest value, or sums of products for weighted means.Browse the other rules of the pandas pack.