Worked examples
- Histogram example: laboratory measurements
- Histogram example: waiting times
- Histogram example: measurement deviations
How to make a histogram
- Paste a column of observations or open an Excel, CSV or TSV file.
- Select the Value column and check missing/invalid exclusions.
- Choose Auto, Number of bins or Bin width; inspect the full bin table.
- Choose Count or Density, adjust size/style, save the project and export.
Data format and interpretation
One raw numeric value per row. Zero and negative values are valid. Keep every repeated observation; do not paste a frequency column as if it were raw data. Tables above 200,000 accepted observations are rejected without sampling.
| Value |
|---|
| 0 |
| 1 |
| 1 |
| 2 |
| 3 |
| 4 |
Auto uses Sturges as a starting choice; different bins can change the appearance. Number mode uses exactly the requested number of bins. Width mode uses a zero-anchored grid unless you supply an origin. All accepted observations remain included. The last bin includes its final edge; other right edges are excluded.
Density is count divided by n and bin width, with reciprocal X units. It does not estimate normality. Exact edges, counts and densities remain in the table even when axis ticks are thinned for readability. Save/open preserves bin choices and source data. Example loading preserves style and offers Undo.
Frequently asked questions
How is a histogram different from a bar chart?
A histogram groups raw numeric observations into adjacent numeric intervals. A bar chart compares categories. This tool does not accept precomputed frequencies as repeated observations.
Which interval receives a value on a boundary?
Bins include the left boundary and exclude the right boundary, except the last bin includes its right boundary. An exact internal boundary goes to the bin on its right. No fuzzy tolerance is applied.
What does density mean?
Probability density is count divided by the accepted number of observations and the bin width. The total bar area equals one; heights need not sum to one and can exceed one.
What happens to missing or identical values?
Missing and invalid values are excluded with source row counts. Repeated values remain repeated observations. A constant dataset receives a finite padded display range in Auto and Number modes.
Can I reproduce the bins in R?
Yes. Download the R script with exact accepted values and realized edges. It uses hist with right=FALSE, include.lowest=TRUE and fuzz=0. You can also download the full bin table.