Task prompt · Data
Find outliers in my data and explain them
by Agent Yard TeamSpot unusual values, check if they are errors or real, and decide what to do with each.
You'll attach
Needs file
Works with
ChatGPT
Claude
Gemini
Grok
CopilotUse cases
- Spot unusual values, check if they are errors or real, and decide what to do with each.
- When you already have a file and want the AI to work from it instead of starting from scratch.
- Reuse it as often as you like: fill in the blanks (column or columns) with new details and copy again.
- As one step of the Make sense of your data prompt pack.
Featured in: Best AI prompts for cleaning messy data.
How to use it
- Open chatgpt.com or claude.ai (or another AI listed here).
- Attach the file (PDF, Word, CSV or Excel), or paste its text into the prompt.
- Fill in the highlighted blank ([COLUMN OR COLUMNS]) below, or edit them after pasting.
- Copy the prompt, paste it into the chat and send.
Fill in the blanks (optional)
Prompt
Look through the attached data for outliers in [COLUMN OR COLUMNS]. 1. Explain the method you used (for example 1.5x IQR or z-score) in one sentence. 2. List each outlier with its row, value and how far it is from normal. 3. For each, say whether it looks like a typo, a unit mix-up, a duplicate or a real extreme value, and why. 4. Recommend keep, fix or remove for each, and show how the main average changes if they are removed. Don't delete anything. Just report.
One email a week
The Weekly is presented byyour brandSponsored · openThe best new prompts and personas, nothing else.