BUG: Fix formatters applied to wrong columns in truncated DataFrame.to_string (GH#35410)

[tinezivic]

• closes BUG: Wrong Custom Formatters applied when displaying trancated frames #35410
• Tests added and passed
• All code checks passed
• Added an entry in the latest doc/source/whatsnew/v3.1.0.rst file
• I have reviewed and followed all the contribution guidelines
• If I used AI to develop this pull request, I prompted it to follow AGENTS.md

Bug

DataFrame.to_string() applied formatters dict to wrong columns when output was horizontally truncated via max_cols.

Minimal reproduction:

import pandas as pd
import numpy as np

df = pd.DataFrame(np.arange(30).reshape(5, 6), columns=[f"Col{i}" for i in range(6)])
formatters = {c: (lambda x, c=c: f"{c}: {x}") for c in df.columns}
print(df.to_string(formatters=formatters, max_cols=4))
# Before fix: Col4 shows "Col2: 4", Col5 shows "Col3: 5"
# After fix:  Col4 shows "Col4: 4", Col5 shows "Col5: 5"

Root Cause

In DataFrameFormatter._get_formatter(), when formatters is a dict and i is an integer (positional index), the code resolved i against self.columns (the original frame). After truncation, positional index 2 in the truncated frame corresponds to e.g. Col4, but self.columns[2] returned Col2.

Fix: resolve positional indices against self.tr_frame.columns (the truncated frame) instead of self.columns. Also removed the i not in self.columns guard which caused incorrect behavior when column names are integers.

AI Disclosure

This fix was developed with the assistance of GitHub Copilot (Claude Sonnet 4.6). The AI assisted with code generation and diff review. The contributor verified the fix locally and confirmed the test passes.

[jbrockmendel]

thanks @tinezivic