DataFrame.applymap was deprecated in pandas 2.1 (August 2023) and emits a FutureWarning. The replacement is DataFrame.map, which does the same elementwise application. Series.map already existed, so the rename makes the two APIs consistent; applymap on a Styler object is unaffected and keeps its name.
The upgrade is mechanical, but there are two behavioural notes worth knowing at the same time. DataFrame.map accepts na_action="ignore", which skips NaN values instead of passing them to your function, avoiding the usual TypeError from a function that expects a number.
More importantly, elementwise mapping over a DataFrame is a Python-level loop over every cell and is orders of magnitude slower than a vectorised expression. Before doing the rename, check whether the call can be replaced entirely: arithmetic, comparisons, .str accessors, np.where, and Series.replace all operate on whole columns in C. Reserve map for genuinely per-cell logic that cannot be expressed any other way.