CSV Tools New Tool

Free Online CSV Statistics and Column Profiler

Profile a CSV column by column in your browser. For every column you get its detected type, how many values are missing, how many are unique, the most common value, and for numbers the minimum, maximum, total, mean, median and standard deviation. The whole profile copies out as a plain text report.

Free Forever Nothing Uploaded Column by column Blanks and uniques

Free Online CSV Statistics and Column Profiler

Advertisement Slot (Top Banner) Google AdSense Unit • Responsive Banner

CSV Statistics Analyzer Your file is read inside this tab. Nothing is uploaded anywhere.

Drop the CSV files you want profiled Nothing is changed. Each file is read and profiled, and the report can be downloaded. Choose files Up to 25 MB per file. Nothing is uploaded.

Or paste it

Reading the file Separator First row is a header Trim spaces around values Dates are day first, 05/01 is 5 January

Show Numeric summary for number columns Most common value in each column Earliest and latest for date columns

Rounding Decimal places Group thousands in the report

0 rows 0 columns 0 empty cells 0 numeric columns

The data

Nothing to show yet.

The profile

0 characters 0 lines

Buy Us A Coffee

Enjoying WizTools123? Help keep our server infrastructure 100% free and open for everyone.

Buy Us A Coffee

Sponsored Content (Below Tool) Google AdSense Placement

How to Analyse a CSV File

A few steps, and nothing is uploaded.

Drop the file in One CSV or several. Nothing is modified; the file is only read and measured.

Read the table One row per column of your file, with its type, how many values are missing and how many are unique.

Look at mean against median When they are close the column is evenly spread. When they are far apart, a few large values are pulling the mean, which is usually worth knowing about.

Copy or download the report The plain text version is laid out for pasting into a ticket, an email or a handover document.

What to Know About These Figures

Which formulas are used, and what the numbers do not tell you.

Mean and median are shown together on purpose. If the average order is 3,400 and the median is 400, the average is being pulled by a handful of very large orders and quoting it alone would be misleading. That gap is often the most useful single fact about a numeric column, and it only shows when both are present.

Standard deviation uses the sample formula, dividing by n minus one. A CSV is almost always a sample of something larger rather than the entire population, and the population formula understates the spread in that case. If you genuinely have the whole population, the figure here is very slightly larger than the textbook one.

Unique counts tell you what a column actually is. A column with as many unique values as rows is an identifier. One with two or three is a flag or a status. One with a few hundred in a thousand rows is a category. That single number usually tells you more about a strange file than the column name does.

Blank counts come before any other question. A column that is forty percent empty cannot support a calculation, no matter what the mean says, and this is the figure to check first on any file somebody else produced.

Numbers are recognised through their formatting. A value with a currency symbol or thousands separators is read as a number for the statistics, while the value itself is never changed. Percentages are read as the figure written rather than divided by a hundred, because that keeps the ordering and the total meaningful.

A column that mixes numbers and text is treated as text. There is no honest average of a column where some rows say 42 and others say "not supplied", so no numeric summary is offered for it. The mixed state is visible in the type, which is the thing worth acting on.

Key Features & Capabilities

What this tool does, and what it deliberately does not.

Every column profiled Type, filled, blank, unique and the most common value, in one table.

Full numeric summary Minimum, maximum, total, mean, median and sample standard deviation.

Mean against median Both side by side, because the gap between them is what reveals outliers.

Date ranges Earliest and latest for date columns, which is how you check an export covers the period you asked for.

A report to paste The whole profile as plain text, laid out for a ticket or a handover note.

Nothing is uploaded The file is measured inside this tab, which is what makes it safe for payroll or customer data.

About the CSV Statistics Analyzer

Somebody sends you a CSV. Before you can do anything with it you need to know what is in it: how many rows, which columns, whether the amount column is really all numbers, how much of it is empty, whether the dates cover the period you asked for. Those questions are quick in a spreadsheet one at a time and tedious all together, and on a file too large to open they are not quick at all.

The figures that matter for that first look are not the ones a statistics package leads with. The blank count comes first, because a column that is half empty cannot support any conclusion. The unique count comes second, because it tells you what the column is: an identifier, a flag, a category or free text. Only then do the averages mean anything.

And when they do, the mean alone is not enough. A single enormous order pulls an average far away from what a typical order looks like, and reporting the mean without the median is how a number ends up in a slide that nobody can reproduce. Showing both costs nothing, and the gap between them is often the first sign that a file needs a closer look.

Frequently Asked Questions

Types, blanks, standard deviation and large files.

Because a few large values are pulling the mean upwards. The median is the middle value and barely moves, so a large gap between them means the column has outliers. In that case the median usually describes a typical row better.

The sample one, dividing by the number of values minus one. That is the right choice when the data is a sample of something larger, which a CSV almost always is.

What kind of column it is. As many unique values as rows means an identifier. Two or three means a flag or status. A few hundred in a thousand rows means a category. It is often more informative than the column name.

Because it is not entirely numeric. One value such as "not supplied" or "n/a" makes the column mixed, and there is no honest average of that. The type shown tells you what happened, and the validator will point at the odd rows.

Yes. A figure written with a currency symbol or thousands separators is read as a number for the statistics, while the value in your file is never changed.

Up to twenty five megabytes, which is roughly two hundred thousand rows. The statistics are computed in one pass, so the limit is about keeping the browser responsive rather than about the arithmetic.