Pandas Iat Float

Pandas Iat Float pulls a single cell out of a DataFrame by row and column position and hands it to you as a Python float. iat is pandas' positional cell accessor, and this is the float version of the extraction family that also includes Pandas Iat Int and Pandas Iat Datetime. It's a tiny node with a quietly big job: it's how you get a number out of your table and into the rest of your graph as a real scalar.

Think about the workflow where a table contains a parameter you need elsewhere. You loaded a CSV of settings, or a summary frame you computed, and somewhere in it is one value - a mean, a max, a coefficient - that should drive something downstream. A conditioning strength, a scale factor, a seed. Without this node you'd be copying the number off a display and typing it into a widget by hand, which is exactly the kind of manual step this pack exists to eliminate. Extract it, wire it in, and the graph stays data-driven.

How it works

dataframe.iat[row_integer_position, column_integer_position] wrapped in float(value). Positional, 0-based: row 0, column 0 is the top-left cell. The cast to float is forgiving - an integer cell, a float cell, even a numeric string will convert - but a text cell that isn't numeric raises a ValueError, and so does an empty/NaN cell. So point it at a numeric cell and you're fine; point it at anything else and the node tells you.

The usual Iat caveats apply: positions are order-sensitive, so a sort, filter, or group-by upstream changes what row N is. If you loaded a CSV and the row order shifts, your extraction silently points at a different value. For label-stable lookups, the Pandas At Float node (row/column by name) is the safer sibling. Use Iat when the frame is stable and you know exactly where things sit.

Inputs and outputs

• dataframe (required, DATAFRAME) - the frame to look into.
• row_integer_position (required, INT, default 0) - the 0-based row.
• column_integer_position (required, INT, default 0) - the 0-based column.
• FLOAT output - the cell's value as a float, wireable into any float-accepting input.

Installing it

Installed with the pack: ComfyUI Manager → search "Data analysis" → install ComfyUI-Data-Analysis → restart → reload. Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/HowToSD/ComfyUI-Data-Analysis
mv ComfyUI-Data-Analysis data-analysis
pip install -r data-analysis/requirements.txt

Requirements: pandas, matplotlib, seaborn, scipy, scikit-learn, openpyxl, lxml. No GPU, no model downloads. If the node errors, check that the cell is numeric - a stray text value or a NaN is the usual culprit, and the fix is data cleaning upstream, not reinstalling anything.