GSoC Sixth Week Progress

Fixing print.data.table() Edge Cases

This week I focused on improving the behavior of print.data.table() by addressing issue #7735.

The issue occurred when col.names="none" was used together with row.names=FALSE. Instead of printing the data without column headers, print.data.table() produced no output at all.

To resolve this, I submitted PR #7804, which changes the internal handling when row names are suppressed. Rather than relying on a pattern that expected row markers, the header is now removed by position, ensuring that the data rows are printed correctly.

Improving .SD Diagnostics

I also worked on issue #7431, which highlighted confusion caused by local assignments to columns that are included in .SD.

Initially, the discussion focused on improving the documentation, but after further feedback the solution evolved into warning users whenever local assignments to .SD columns are detected.

I submitted PR #7807, which includes:

• A recursive scanner to detect local writes to columns referenced by .SD while avoiding false positives from read operations.
• Documentation updates explaining the behavior of .SD and the new warning.
• Regression tests covering grouped, ungrouped, and false-positive scenarios.

This work helps make .SD behavior more explicit while preserving existing workflows.

Looking Ahead

With these improvements completed, I’ll continue working on documentation and consistency improvements across data.table, while following up on review comments for the open pull requests. ```