Number List Sum

Features

• Count, total, mean, median, min, max, range, sample variance and sample standard deviation in one pass
• Plus mode and quartiles (Q1, Q3, interquartile range) — every tied mode is listed, and “no mode” is said plainly when no value repeats
• Two ways to take numbers: automatic extraction, or split by a separator
• Automatic mode understands thousands separators, decimals, scientific notation and signs
• How many items were skipped, and which ones, are both listed with a few examples
• Adjustable precision, with trailing zeros dropped, so 1 + 2 shows as 3 rather than 3.000000
• Skipped items count towards neither the total nor the item count
• Raises an error when there are no numbers at all instead of a table of zeros
• The result is an aligned table, and every row can be copied on its own
• Everything runs locally in your browser; nothing is uploaded

How to use

1. Paste the numbers (one per line, or separated by commas or spaces)
2. Choose split by a separator, or automatic extraction
3. The result updates as you type; anything skipped is listed separately
4. Copy a single row, or copy the whole table

FAQ

Why is the total different from what I worked out myself?

First look for a Skipped row. If it is there, some items could not be parsed as numbers — things like `12kg`, `N/A` or `50%` — and they were left out of the total; that row exists to tell you so. If you do want them counted, remove the units and symbols, or switch to automatic extraction, which will pull the 12 out of `12kg`.

What does something like `3-5` count as?

In automatic mode it becomes 3 and 5 — two items — because a minus sign immediately after a digit is read as a range separator rather than a sign. That is deliberate: in running text `3-5` is almost always a range, while ` -5` (after a space or at the start of a line) is read as minus five. For an exact reading, switch to splitting by a separator, where `-5` is always minus five.

Is the standard deviation for a population or a sample?

It is the sample standard deviation (dividing by n−1). The reason for that choice is that the input is nearly always a batch of observations — a day of takings, a run of part dimensions — and what you want to know is how much the thing behind them varies, which calls for the sample estimate of the population. With a single number the denominator would be zero, so that cell is left blank rather than filled with an invented 0.