Pandas Show Series

You're three nodes into a data pipeline, the output type is a mysterious "PDSERIES," and all you want is to look at the thing. This is the node for exactly that. Pandas Show Series takes a Series - the one-column pandas object that most summary nodes in this pack spit out - and renders it as text right inside the node's widget. It's a debugging display node, not a transformer. It doesn't change your data; it shows you what your data became.

How it works

This pack passes real pandas objects across the wires between nodes, but ComfyUI's UI has no idea how to draw a Series. So the pack borrowed the classic trick from pythongosssss's show_text node (the source file credits it) to get the value onto the canvas: on execution it serializes the Series to a JSON string holding its values, index, and name, then writes that string back into its own widget by finding its node ID in the workflow JSON. Two nice consequences: the text updates after every run, and because it's stored in the widget, the value gets baked into the saved workflow file. Share a .png of your workflow and whoever opens it sees the data you saw.

Inputs and outputs

One input that matters:

• series_list (PDSERIES) - wire any Series output into this. Despite the plural name, just connect a single Series; the node declares INPUT_IS_LIST, so ComfyUI wraps it for you.

The single output is STRING (as a list), which is the same JSON text shown on the widget. You can chain it into a text node if you want, but honestly the widget is the destination.

Typical wiring: Pandas Sum, Pandas Std, Pandas Mean, Pandas Skew, or Pandas Select Column As Series → Pandas Show Series. Run the graph, resize the node, and read the numbers.

One honest gotcha: what you get is JSON, not a tidy table. {"array": [10, 12, 15], "index": [0, 1, 2], "name": "hits"} is informative but ugly. If you want a readable grid instead, feed a DataFrame to Pandas Show DataFrame, or use Pandas To String into Pandas Show Text.

Installing this pack

This node ships in ComfyUI-Data-Analysis by Hide Inada (HowToSD). No GPU and no model downloads - it's pure CPU pandas work, which is the point. The catch: pandas and friends aren't in a stock ComfyUI, so the pack pulls in its own requirements (pandas, matplotlib, seaborn, scipy, scikit-learn, openpyxl, lxml).

ComfyUI Manager (easiest): Manager → Custom Node Manager → search "Data analysis" → install ComfyUI-Data-Analysis → restart ComfyUI and refresh the browser. Manager handles the Python deps.

Git clone:

cd ComfyUI/custom_nodes
git clone https://github.com/HowToSD/ComfyUI-Data-Analysis.git
mv ComfyUI-Data-Analysis data-analysis   # the rename matters, see below
pip install -r data-analysis/requirements.txt

The mv isn't optional cosmetics: the pack's example workflows reference the data-analysis folder name, so skipping it breaks them. Restart ComfyUI after installing.

Troubleshooting

• The widget is empty after a run. The text is only written when the node actually executes, so an output node that's sitting unconnected or bypassed shows nothing. Check the wire is attached, then queue again.
• It says "PDSERIES" but the wire won't connect. Series inputs only accept Series outputs. If you've got a DataFrame, use Pandas Select Column As Series to pull one column out first, or use Pandas Show DataFrame instead.
• This is an output node. It has no downstream effect, so don't expect its string to change anything further along the graph.