Free CSV tool
How to group by and aggregate a CSV file
Roll a CSV up by one or more columns and compute sum, mean, median, count, count distinct â or a proper 95% confidence interval on the mean, not just a point estimate. All of it in your browser, on the actual data, not a sample.
Runs entirely in your browser â your file stays on your machine.
How it works
- Load your file. One CSV/TSV/XLSX â grouping and aggregating happens within a single table.
- Add an aggregate step (already set up if you arrived here from this page) and pick one or more group-by columns.
- Add a metric for the column you want to summarise, and check the aggregations you want: sum, count, count distinct, mean, median, or a 95% CI.
- Choose how the CI displays â as separate lower/upper columns, or one formatted "12.34 ± 1.56" column â if you selected the confidence interval.
- Apply and see the grouped, aggregated table update immediately.
Worked example
Revenue by region, with a confidence interval
Group a sales file by region, aggregate revenue with sum, mean, and the 95% CI:
region | revenue_sum | revenue_mean | revenue_ci95_lower | revenue_ci95_upper east | 48200 | 1205.00 | 1041.32 | 1368.68 west | 61900 | 1547.50 | 1298.11 | 1796.89
The interval comes from a Student's t-distribution over the actual sample size in each group (not a fixed 1.96Ã), using Bessel-corrected sample variance â the same defensible method a researcher would use by hand, computed instantly per group.
Count vs. count distinct
Plain count counts every non-empty value in that column within the group. count distinct counts unique values â useful for, say, "how many distinct customers ordered in each region" versus "how many orders were placed." Both are exact, not sampled or estimated.
Related tools
Aggregate your data now
No signup, no upload â the workspace opens with a group-by step ready.
Open the workspace