Line Chart Maker

Paste a header row and a category column and every remaining column becomes a line series. The chart renders with Apache ECharts in the page, so nothing is uploaded, and it redraws as you edit the data rather than on a submit button.

How to use it

1. Paste CSV or TSV directly from Excel, Sheets, or a query result.
2. Confirm the first column is being read as the X axis and each remaining column as a series.
3. Edit values in the grid and watch the chart redraw.
4. Pick a theme and download a 2× resolution PNG.

The data layout it expects

The parser wants wide format: one row per point on the X axis, one column per series. A header row is required, because the column headers become the series names in the legend.

So a table with Month, Revenue, and Profit columns produces two lines across twelve months. If your data is in long format instead — one row per category-and-series pair, with the series name in its own column — you either need to pivot it first or use the combo chart maker, which reads long format natively.

Pasting straight from a spreadsheet works because the clipboard carries tab-separated values, and tabs are detected alongside commas. That matters more than it sounds: it means you do not have to export a file, and since nothing is uploaded, the data never lands on disk anywhere in the process.

Values are parsed as numbers where possible. Currency symbols, thousands separators, and percentage signs are common in spreadsheet exports and will not parse cleanly, so strip them in the source column before pasting. Empty cells are treated as gaps rather than zeros, which is the correct default — a missing measurement and a measurement of zero are different claims, and drawing a missing month as a plunge to zero is a real distortion.

What a line implies that a bar does not

Connecting two points with a line asserts that the values in between are meaningful. That assertion is true for a quantity measured over time and false for separate categories, which is the single most common misuse of this chart type.

Revenue by month is continuous: the line between March and April represents real intermediate values you did not sample. Revenue by product line is not: there is nothing between Product A and Product B, and drawing a slope there implies a trend that does not exist. Use bars for categories.

The same logic governs the X axis spacing. This chart treats the category column as evenly spaced labels, not as dates on a true time axis. If your rows are at regular intervals — every month, every day — that is correct and simpler. If your intervals are irregular, with three readings in January and one in June, even spacing will visually flatten the gap and misrepresent the rate of change. In that case add the missing periods as empty rows so the spacing carries the right information.

Axis scaling and the honesty of the baseline

The Y axis is auto-scaled to the data range rather than pinned to zero, which is deliberate and worth understanding, because it is also the most frequent complaint about line charts.

For a series moving between 98 and 102, forcing a zero baseline compresses every real change into a flat band and hides the signal entirely. For a series moving from 20 to 80, a truncated axis exaggerates the change dramatically. Neither default is universally honest.

The convention that holds up is that line charts showing change over time may truncate the axis, because the reader is being asked about the shape of the trend, while bar charts must start at zero, because the reader is being asked to compare the areas of the bars. A truncated bar chart misrepresents the ratio between two values; a truncated line chart does not make the same claim.

Where the axis is truncated and the difference matters, label it. Readers who notice a non-zero baseline distrust the chart, and readers who do not notice may take away a stronger conclusion than the data supports.

How many lines is too many

Legibility falls off sharply above about five or six series. Beyond that the lines cross often enough that following any individual one requires constant reference to the legend, and colour differences stop being reliable — particularly for the roughly 8 percent of men with a red-green colour vision deficiency.

Two approaches work better than adding more colours. Highlight one or two series in strong colours and render the rest in a light grey, which keeps the context visible while making the comparison obvious. Or split into small multiples: several small charts on a shared scale, one series each, which the eye can scan far faster than a tangle of lines.

Direct labelling at the end of each line also outperforms a legend, because it removes the lookup step. With few enough series that the labels do not collide, it is almost always the better choice.

Exporting for print and slides

The PNG export renders at twice the display resolution, so a chart occupying 800 by 400 pixels on screen exports at 1600 by 800. That is adequate for slides and screen documents and marginal for print at 300 DPI, where 1600 pixels covers about 5.3 inches.

The Print theme exists for that case: darker strokes, higher contrast, and a light background, since dark-background charts consume a great deal of toner and lose contrast on paper. Switch themes before exporting rather than after, since the export captures the current render.

Font sizes are set relative to the chart, so a chart that reads well on screen will usually read well when scaled. The common failure is exporting a wide chart and then placing it small in a document, at which point axis labels drop below legible size. Sizing the chart close to its final dimensions before export avoids this.

At a glance

Frequently asked questions

Read more

Charting change over time — Connecting two points asserts that the values between them exist. That assertion is the whole difference between a line chart and a bar chart.

Related chart makers