# Weight Chart Data Extraction Checklist

> Turn a hand-drawn or printed weight chart into date and weight pairs. Read the axes, read every point, compare the labels and keep the units.

## What this checklist is for

A weight chart on paper, in a scan or in a photo of a screen contains data. You cannot sort, average or graph that data until each point is a date and a number. Errors in chart reading are predictable. People misread the axis scale, skip points or mix up pounds and kilograms. A written label can also disagree with the position of its point. This checklist turns any weight chart, hand-drawn or printed, into a list of date and weight pairs.

For each pair, you also record how sure you are. The checklist works for the weight log of a person, the growth chart of a child or the weight record of a pet. Only a qualified clinician can make a health decision from the numbers. For an animal, ask a veterinarian.

## Before you read the points

1. Read the title and the units. Write down whose weight the chart shows and the unit, lb or kg. Also write down what the horizontal axis shows, for example dates, months, weeks or age.
2. Look at the vertical axis. Make sure that the tick labels go up in equal steps. A hand-drawn chart sometimes has a wrong tick label, for example the same number two times. If the labels do not go up in equal steps, write that down before you read a value.
3. Look at the horizontal axis. A hand-drawn chart can have uneven space between dates. Read each date from its own label. Do not count grid squares.
4. On a growth chart, read only the plotted points. The printed curves on the growth chart of a child are percentile lines. A percentile line shows values for a reference group of children. These lines are not measurements of the child. The CDC [recommends](https://www.cdc.gov/growth-chart-training/hcp/overview/recommended.html) the WHO growth standard charts from birth to 2 years. For ages 2 to 20, it recommends the CDC growth charts. Write down which chart you read.

## The checklist

1. Count the points first. Write down the number of plotted points. The finished table must have the same number of rows.
2. Read the date of each point from the nearest axis label. If a point is between two labels, estimate the date. Mark that date "estimated".
3. If the chart has a number next to a point, record that number in its own column. Do not assume that the number is correct.
4. Read the position of each point against the vertical scale. Read to the precision of the grid. Record this value in a separate column.
5. Compare the written number with the position. If they differ by more than one grid step, flag the point. Do not choose one value without a note.
6. Ignore arrowheads, line ends and the start of an axis. They are not points unless the chart marks them as points.
7. Keep the original unit. If you need another unit, convert in a separate column. Convert from the value before you round it. NIST gives the [conversion factor](https://www.nist.gov/pml/special-publication-811/nist-guide-si-appendix-b-conversion-factors/nist-guide-si-appendix-b8) 1 lb = 0.4535924 kg.
8. Make sure that the dates are in order and that no date appears two times. If two neighboring points have a large jump, read both points again.
9. Write down what is missing. Examples are an axis without a unit, dates without a year and points outside the chart. Also write down handwriting that you cannot read.

## How to record the result

Use one row for each point:

| # | Date as shown | Full date (if known) | Weight from label | Weight read from axis | Unit | Estimated? | Flag | Notes |
|---|---|---|---|---|---|---|---|---|
| 1 | | | | | | | | |

Under the table, write the unit of the chart and the grid step that you read to, for example the nearest 1 lb. Also write the number of points that you counted and the number of rows. If the two numbers match, copy the date column and the weight column into a spreadsheet.

## Use this checklist with Claude or ChatGPT

You can attach a photo or a scan of the chart to Claude or ChatGPT, together with the checklist below. The prompt asks for the written number and the axis reading of each point. Then you can see each difference between the two.

```text
You are helping me extract data points from a weight chart image.

Checklist:
1. Read the chart title, the weight unit and what the horizontal axis shows.
2. Check that the vertical axis labels go up in equal steps. Report any label that breaks the pattern.
3. Count the plotted points and report the number before you list them.
4. For each point, give the date from the horizontal axis. If it falls between labels, estimate it and mark it "estimated".
5. For each point, give the number written next to it (if any) and, separately, the value you read from its position on the vertical axis.
6. Flag any point where the written label and the axis reading differ by more than one grid step.
7. Do not treat arrowheads or line ends as data points unless they are marked as points.
8. Keep the chart's unit. Do not convert.

Output:
- A table with the columns: number, date as shown, weight from label, weight read from axis, unit, estimated (yes or no), flag.
- For each checklist item, one or two lines on how you checked it.
- A list of everything you could not determine, and why.

Chart:
[attach the image]
```

Before you use the result:

- Compare the number of points with the chart. Read several values yourself. AI tools can skip points, misread handwriting or correct a mismatch without a note.
- Crop out names, record numbers and other personal details before you upload. Also read your workspace's data settings.
- Do not use the numbers as an assessment. For a person or a child, a qualified clinician must review what the numbers mean. For a pet, ask a veterinarian.

A docs portal with Ask AI
and support tickets

Connect a repository, or describe your product to start. Each docs site has search, Ask AI, an API console, an MCP server and support tickets.

Publishes from GitHub on every push

Works with Mintlify repositories

API console from your OpenAPI file

MCP server and llms.txt for AI agents

Ask AI and support tickets

Analytics for people and AI agents

Start Free →

Questions? Ask AI in our docs
No card required.

---

[View the canonical page](https://revdoku.com/blog/weight-chart-reading/) · [Browse llms.txt](https://revdoku.com/llms.txt)