Summary statistics for every numeric column in a CSV

How to use the summary statistics calculator

1. Press Choose a file to load a CSV, TSV or Excel file, or paste your table into CSV data.
2. Check the Delimiter and keep First row is a header ticked if the first row has column names.
3. Choose Variance and standard deviation: Sample (n − 1) or Population (n).
4. Set Decimals (0–10) for the results.
5. Read the Summary statistics table, one column per numeric field, and use its CSV button to save summary-statistics.csv.

What it does and when to use it

Before you chart or report on a data set, it helps to know its centre and spread. This tool gives a first look at every numeric column at once: test scores per subject, prices in an order export, response times in a log, or monthly figures in a budget.

The mean and median together tell you whether a few large values pull the average up. The standard deviation and interquartile range show how spread out the values are. Quartiles are also the numbers behind a box plot. To see how often each value appears, use the frequency counter. For totals by group, use the pivot table builder. For a single list of numbers typed by hand, the mean, median and mode calculator is simpler.

How it works

Each numeric column is treated as a list of n numbers x, with empty cells left out:

mean      = Σx ÷ n
variance  = Σ(x − mean)² ÷ (n − 1)    sample
variance  = Σ(x − mean)² ÷ n          population
std. dev. = √variance
range     = maximum − minimum

For the median (p = 0.5) and quartiles (p = 0.25 and 0.75) the values are sorted and the position h = (n − 1) × p is found. If h falls between two values, the result is the value at the lower position plus the fraction of the gap to the next one. The interquartile range is Q3 − Q1. A sample variance needs at least two values, so a column with one number shows “—”. Everything runs in your browser.

Worked examples

Class results (the sample). Eight students with maths, science and attendance columns; the student names are skipped as text. Part of the table:

Statistic | maths | science | attendance %
Mean | 75.88 | 79.88 | 90.38
Median | 78 | 80.5 | 92.5
Mode | 78 (×2) | none (no repeats) | none (no repeats)
Variance (sample) | 149.55 | 127.27 | 78.55
Standard deviation | 12.23 | 11.28 | 8.86

The textbook set 2, 4, 4, 4, 5, 5, 7, 9. With Population, the mean is 5, variance 4 and standard deviation 2, with Q1 4, median 4.5 and Q3 5.5. With Sample the variance is 4.57 and the standard deviation 2.14.

Messy prices. $4.50, "$1,200", (45) and N/A with 1 decimal. Three values are read (4.5, 1,200 and −45) and N/A is counted under Non-numeric cells ignored | 1. The mean is 386.5 but the median only 4.5, a sign that one large value is pulling the average up.

Limits and tips

• Decimal commas such as 1,5 are not read as 1.5; switch them to points first.
• Dates are not numbers here, so a date column is skipped.
• Spreadsheet files use their first sheet only.