Visualize a CSV as a chart
To chart a CSV, open the chart builder in Sigmera, pick the category to group by and the value to summarize (sum, average, count and more), choose a bar, line, area or pie chart, and export a PNG or SVG. It all happens in your browser — your data is never uploaded to a server, so it’s GDPR-safe by design.
Last updated: September 2026
Use the chart builder free in your workspace
Free account, no credit card — every tool, unlimited rows, 3 free downloads a month, nothing uploaded.
🔒 Everything runs in your browser — your file is never uploaded, and the workspace even works offline once it has loaded.
How to make a chart from a CSV
- 1. Open the chart builder in your free workspace. Your file is read into the browser’s memory — never sent over the network.
- 2. Group and summarize. Pick the category column for the X axis and how to summarize the value — sum, average, count, min, or max — optionally split into series.
- 3. Choose a chart type. Switch between bar, line, area, and pie and watch it update live.
- 4. Export the image. Download a PNG or SVG with your title and data baked in.
A CSV plotter skips the spreadsheet step
A CSV plotter reads the raw text file and draws the chart directly, rather than importing the rows into a workbook first. That removes three steps most people take by reflex: opening the file in Excel, fixing the columns the import mangled, and building a pivot table before the chart wizard has anything to draw from.
It matters most on the files you did not create — a platform export, a survey download, a monthly billing extract. You want one number out of it: revenue by region, tickets by month, signups by source. Going CSV to chart directly gets you that number in about four clicks, and you never touch the delimiter and encoding problems that come with opening a CSV in a spreadsheet.
The chart builder wants the file tall: one row per observation, a column to group by and a column to summarize. Exports do not always arrive that way — a billing extract with one column per month is wide, and there is no pair of columns in it to chart. Reshaping that file before you load it is a spreadsheet job, and rotating rows into columns walks through both routes — a straight transpose when the whole table was written the wrong way round, and an unpivot when the headers are really values.
One caveat: a chart is only as honest as the rows behind it. Duplicate records inflate every sum, and the same category spelled three ways splits one bar into three. Run duplicate removal over the file first, and if the text columns are inconsistent, the walkthrough on cleaning data before you report on it covers the rest. When you need the chart inside a workbook instead of as an image, convert the cleaned file with the CSV to Excel converter and chart it there — the rest of the in-browser toolset works the same way, with nothing leaving your device.
Sometimes the source is not a tidy CSV yet. If the facts are spread across a folder of notes, logs, and data files, Lapu AI can search and extract data across local files before you clean or chart the structured result here.
The other boundary appears after the chart. For a recurring report that also has to open a spreadsheet, run a script, and place the finished chart in a presentation, see how Lapu AI handles multi-step workflows across desktop apps.
Which chart type fits your CSV
The chart type is not a style choice — it decides what a reader can actually compare. Microsoft’s guidance on available chart types makes a reasonable default: a column or bar chart for categories with no natural order (bars when the labels are long), a line chart for values at equal intervals such as months or quarters, an area chart when the total across a trend matters as much as the parts, and a pie chart only when you have a single series, no negative values, and no more than seven categories.
Most CSV exports land in the first two. A platform export grouped by source, campaign, or region has no inherent order, so bars read cleanest — and when the category names are long, a horizontal bar keeps the labels from stacking at an angle. A billing or signup extract with one row per date is a line, and there the summarize step matters more than the chart type: counting rows per month answers a different question than summing a value column over the same months.
The pie constraint is the one broken most often. A file with forty categories in it produces forty slices, most of them a sliver, and the chart says less than a bar chart sorted by value would. Charts are also the clearest example of what a CSV cannot carry: the file holds values, not a picture of them, which is part of what separates a CSV from a workbook. Going CSV to chart in the browser gives you the image without making the workbook the middleman.
When a CSV plotter draws the wrong chart
Most bad charts are bad data, not a bad chart type. Two problems account for nearly all of them, and both are visible in the file before you plot anything.
The first is a value column that is text. A CSV has no types — every field is characters — so 1,240, $1,240 and 1 240 are strings, not numbers. Summing them gives you nothing to add, and the chart flattens to a row of identical bars because the only thing left to measure is how many rows there are. Removing the currency symbol and the thousands separator from that one column fixes it; a column-scoped find and replace does it without touching the labels elsewhere in the file.
The second is date ordering. Group a CSV by a date column and the categories sort as text, so April precedes January and 10/03 precedes 2/03. An ISO-style value — 2026-01, 2026-02 — sorts correctly by the same alphabetical rule, which is why reformatting the date column is worth doing before you go from CSV to chart rather than fighting the axis afterwards. It is the same reason exports are worth normalizing in general: the pre-import checklist catches both of these before they reach anything downstream.
Neither fix requires a spreadsheet, and neither requires sending the file anywhere. You visualize the CSV in the same tab you cleaned it in, and the data never leaves the machine it started on.
Sum, average or count — the summary decides the story
Grouping a CSV collapses many rows into one bar, and the function you pick decides what that bar means. The chart looks the same either way, which is what makes a wrong choice hard to spot: five bars of plausible height, measuring something nobody asked for.
Sum answers “how much in total” and only works on quantities that add up — revenue, units, hours. It is wrong on anything already expressed as a rate. Summing a conversion-rate column gives 340%, and summing a per-unit price gives a number that corresponds to no transaction. Average answers “how much typically”, and comes with its own trap: an average of per-region averages is not the overall average unless every region has the same row count. If the groups are different sizes, sum the numerator and the denominator separately and divide those.
Count ignores the value column entirely and measures rows, which is the right answer more often than people expect — tickets per month, signups per source, orders per rep. It is also the summary a duplicate row distorts most directly, since a repeated record is a whole extra unit of whatever you are counting. Clearing those first is why removing duplicate rows before you summarize changes the chart more than any styling choice will.
Blanks are the last thing to check, because Sum and Average disagree about them. Microsoft’s definition of the spreadsheet equivalent is explicit: “if a range or cell reference argument contains text, logical values, or empty cells, those values are ignored; however, cells with the value zero are included” (Microsoft, AVERAGE function). So a column where missing readings were exported as 0 pulls the average down, while the same column with those cells left empty does not — two charts, two different conclusions, from one export. Decide which one a blank means before you plot it; that judgement is part of what data cleaning actually involves, not a charting setting.
Frequently asked questions
- Does this tool upload my data?
- No. The chart is built entirely inside your web browser using client-side JavaScript. Your file is never uploaded to a server, so it is GDPR-safe by design.
- What chart types are supported?
- Bar, line, area, and pie. You group by a category column and summarize a value column with Sum, Average, Count, Minimum, or Maximum — and you can break the data down into multiple series.
- Can I export the chart?
- Yes. You can export the chart as a PNG (for slides and docs) or an SVG (for crisp, editable vector graphics), with the title and your data baked in.
- Does it work with Excel files too?
- Yes. You can chart a .csv, .tsv, .xlsx or .xls file. Sigmera auto-detects the encoding and lets you pick the sheet for multi-tab workbooks.
- What is a CSV plotter?
- A CSV plotter reads a comma-separated file and draws a chart from two of its columns — one to group by, one to summarize. It skips the spreadsheet step: no importing, no pivot table, no chart wizard. Sigmera's CSV plotter runs in the browser tab, so the file stays on your machine.
- How do I turn a CSV into a chart without Excel?
- Open the chart builder, drop the file in, and pick the category and value columns. You do not need Excel, Google Sheets, or any install — the parsing and the drawing both happen in client-side JavaScript. That also means it works on a machine where you cannot install software.
- How large a CSV can it plot?
- There is no server-side row cap, because there is no server. The practical limit is your device's memory — a file with hundreds of thousands of rows charts fine on a normal laptop, since the grouping happens once and the chart draws from the summarized totals, not the raw rows.
- Which chart type should I use for my data?
- Bar or column for categories with no natural order — bars when the labels are long. Line for values at equal intervals such as months or quarters. Area when the total across the trend matters as much as the individual parts. Pie only for a single series with no negative values and no more than seven categories; that limit comes from Microsoft's own chart-type guidance.
- Why does the same category appear twice in my chart?
- Because the grouping is exact. "Acme Ltd" and "acme ltd" are two different keys, so they become two bars, and a trailing space does the same thing invisibly. Normalizing the category column before you chart it collapses them back into one.
- Can I plot a CSV on a locked-down work laptop?
- Yes. There is nothing to install — the chart builder is a web page, and both the parsing and the drawing run in the browser tab. That makes it usable where IT blocks desktop installs but allows the browser.
- My CSV plotter shows every value as 1 — what went wrong?
- The value column is being counted rather than summed, or it is text rather than a number. A column exported as "1,240", "$1,240" or "1 240" is a string, so a Sum has nothing numeric to add. Strip the thousands separator and the currency symbol from that column first, then summarize it again.
- Why are my months out of order when I plot a CSV?
- Because a CSV stores dates as text, and text sorts alphabetically — so "April" lands before "January" and "10/03" before "2/03". Grouping on an ISO-style date column (2026-01, 2026-02) sorts correctly on its own, which is why an exported date column is worth reformatting before you chart it rather than after.
- Should I clean the CSV before charting it?
- Usually yes. Duplicate rows inflate a sum, and inconsistent category spelling ("UK", "uk", "U.K.") splits one bar into three. Removing duplicates and normalizing case first is what makes the chart trustworthy.
- Should I sum or average this column?
- Sum quantities that add up — revenue, units, hours. Average things measured per record, such as order value or response time. Never sum a column that is already a rate: adding conversion percentages together produces 340% and means nothing. And be careful averaging averages — the mean of five regional averages only equals the overall average if every region has the same number of rows. When the groups are different sizes, sum the numerator and the denominator separately and divide.
- How are blank cells handled when I chart a CSV?
- A blank is skipped by an average but is not the same as a zero, and that difference moves the chart. Microsoft's AVERAGE documentation puts it plainly: empty cells are ignored, while cells with the value zero are included. So an export that writes missing readings as 0 will show a lower average than the same export that leaves them empty. Decide what a blank means in your data — genuinely zero, or not recorded — before you summarize the column. Blanks in the category column are different again: they group together as their own bar.