Pandas Rename

Every real dataset has at least one column name you want to strangle - a typo, a "value " with a trailing space, or something imported from a CSV with a name that makes no sense. Pandas Rename is the node for exactly one rename at a time: pick the axis, type the old label, type the new label, done. It's the "clean up your data before you do anything else" node, and in a pipeline it usually sits right after a load or a create.

It's part of HowToSD/ComfyUI-Data-Analysis, the extension that turns pandas operations into ComfyUI nodes. Rename is listed under Transformation, and there's a deliberate pair here: this node handles a single label with plain text fields, while Pandas Rename Advanced takes a whole dictionary of renames at once. For the one-off cleanup, this is the one you want - no syntax to get wrong.

How it works

The node calls pandas' DataFrame.rename() with a single-entry mapping. The axis dropdown picks which label namespace you're editing: columns renames a column header, index renames a row label. Then current_label is the exact label as it exists today and new_label is what it becomes. Case matters, spaces matter - if you don't hit the label exactly, pandas doesn't rename anything and silently hands you back the same table.

One honest limitation the author flags in the source: renaming works for string labels. If your columns or index are integers, this node's plain-string approach isn't the right tool - go for Rename Advanced, which can handle numeric keys.

Inputs

• dataframe - the table.
• axis - dropdown: columns or index.
• current_label - the existing label, exactly as it appears.
• new_label - what to rename it to.

Output is the updated DATAFRAME, everything else untouched.

Installation

Standard pack install. ComfyUI Manager: search Data analysis, install ComfyUI-Data-Analysis, restart, reload. Manually:

cd ComfyUI/custom_nodes
git clone https://github.com/HowToSD/ComfyUI-Data-Analysis.git
mv ComfyUI-Data-Analysis data-analysis   # example workflows rely on this name
pip install -r requirements.txt

Dependencies: pandas, matplotlib, seaborn, scipy, scikit-learn, openpyxl, lxml. No GPU, no models.

The gotcha

Silent no-ops. Because pandas rename just doesn't match when the label is slightly off, the node won't error - you'll think you fixed the column and it'll still be there. Before you go looking for a bug downstream, check that the current_label is byte-for-byte identical (trailing spaces from CSV imports are the classic culprit). And remember: one rename per node instance. If you're fixing five columns, either chain five of these or jump straight to Pandas Rename Advanced, which does them all in one dict.