Line Graph Maker & Line Graph Generator
Make line graphs of time courses, growth curves and dose-response data, with every point, error bar and axis tick computed from your numbers before the figure is drawn. For lab reports, theses and journal panels.
Line graph styles
Five journal presets, each on its own study: Per2 and Bmal1 cycling in antiphase over a day, hypocotyl growth in Col-0 and phyB-9 under red light, MC38 tumour volume under three treatments, HbA1c change from baseline over 52 weeks and an amylase's temperature optimum. Take a card to load its data.
Line graph examples
Four line graph examples, drawn by this tool from the prompt and numbers printed on each card: an E. coli growth curve in LB and M9 glucose on a log₁₀ OD600 axis, HeLa viability under cisplatin and doxorubicin on a log concentration axis, a glucose tolerance test in wild-type and db/db mice with mean ± SEM worked out from four mice per time point, and systolic blood pressure in three trial arms at weeks 0 to 24, spaced as measured. Click one to load its prompt and numbers into the composer.
What is a line graph?
A line graph plots a measured value against an ordered x variable, usually time or dose, and joins successive points with line segments so the reader sees how the value changes. Each line is one condition or group. Because the x axis is a number line, or an ordered set of visits, the spacing between points carries meaning, and the slope between two points is the rate of change over that interval. In research figures the points are usually means, each with an error bar.
About this line graph creator
- Type the x values once (Time (h): 0, 2, 4, 8) and one line per condition, or attach a CSV, TSV or Excel table with x in the first column and one column per line, up to 6 lines and 40 x values.
- Columns that share a header are read as replicates and averaged in code. Error bars are SD by default, SEM or a 95% CI when you name them, and a column headed like “KO SD” becomes that line's ± values.
- Uneven sampling times keep their true spacing on the x axis. Ask for a log x axis for concentrations or a log y axis for counts and OD, and the ticks land on powers of ten.
- Each line gets its own marker shape, the legend lists the lines in the order they end on the right, and it goes in a corner of the plot the data leave empty.
- Every regeneration rebuilds the points from your data; a new journal preset or palette restyles the lines and leaves each point where it was.
Line graphs you can make
The six line graphs that come up most in lab work and coursework: what makes each one different, and what to type to get it.
Line graph over time (time series)
One measurement followed through time, one line per condition. Use it for cell counts over days, body weight over weeks or a signal across a recording. Give the time points as numbers (Day: 0, 3, 7, 14, 28) and the gaps stay proportional, so uneven sampling is not disguised as even. Put the unit in the x title: Time (h), Days after treatment.
Line graph with error bars
Each point is a mean, and its bar shows the spread or the uncertainty around it. Type replicates per time point as 1.0/1.2/0.9, or attach replicate columns that share a header, and the means are computed in code. SD describes how much the replicates scatter; SEM and the 95% confidence interval say how well the mean itself has been pinned down. Name the one you want and state it, with n, in the legend.
Double line graph and multiple line graph
Two to six lines on one pair of axes, compared over the same x values. A double line graph sets treated beside control; more lines suit dose groups or genotypes. Keep only the lines a reader must compare: past four or five the colours and markers start to blur, and stacked panels on a shared x axis read better.
Bacterial growth curve on a log scale
Cell density or viable count against time on a logarithmic y axis, where exponential growth becomes a straight line. On this semi-log plot the lag, exponential, stationary and death phases separate clearly, and the slope of the straight stretch gives the specific growth rate. Ask for a “log scale y-axis” and give OD600 or CFU/mL values above zero.
Dose-response curve
Response against concentration on a log x axis, from no effect to the full effect. Concentrations usually rise in tenfold or threefold steps, which a log axis spaces evenly. Write “log x axis”; a vehicle dose of 0 has no place on a log axis and is left off with a note. The measured points are joined as given: no sigmoid is fitted and no EC50 is drawn that you have not supplied.
Repeated measures from baseline to follow-up
Group means at scheduled visits, often as change from baseline. Clinical and intervention studies report Baseline, Week 4, Week 12, Week 24. Give the weeks as numbers to keep the real spacing, or as labels to space the visits evenly. Lines that start at 0 and fall below it show the change directly, and the y axis runs below zero to hold them.
How to make a line graph
Enter x and the lines
Type the x values once, Time (min): 0, 15, 30, 60, then one line per group, WT: 98, 208, 178, 142, or attach the plate-reader or spreadsheet export as CSV or Excel.
Say what the bars and axes show
Name the error bar (SD, SEM or 95% CI), ask for a log axis when the values span decades, and give each axis title with its unit in quotes.
Pick a journal style and generate
Pick the Nature, Science or Cell Press preset, clinical grayscale, or a dark slide for talks. The points, ticks and legend order are fixed by your data, so a preset changes only line weight, markers and colour.
Choosing the axes and error bars for your data
Most line graph mistakes are axis choices: evenly spaced ticks for uneven times, a linear axis for values that grow a thousandfold, a second y axis. Match your data to the setting.
| Your data | X axis | Y axis | Error bars |
|---|---|---|---|
| Samples at 0, 1, 2, 4, 8 and 24 h | Linear time at the true spacing | Linear | SD or SEM of independent repeats |
| OD600 or CFU/mL during growth | Linear time | Log₁₀ (semi-log plot) | SD of biological replicates |
| Response at 0.1 to 10,000 nM | Log₁₀ concentration | Linear, % of control | SEM or 95% CI |
| Visits at baseline and weeks 4, 12, 24 | Weeks as numbers, or visits as evenly spaced labels | Change from baseline, crossing 0 | 95% CI |
| Two variables in different units | One shared time axis | Two stacked panels, one y axis each | As for each variable |
| One representative experiment | As above | As above | None, since n = 1 |
Error bars belong to independent repeats, not technical replicates. A single representative run has n = 1 and carries no error bars, and replicate wells from one run show pipetting precision rather than biological variation (Cumming, Fidler and Vaux, 2007).
Slope, growth rate and doubling time from a line graph
The slope between two points of a line graph is the average rate of change over that interval; on a semi-log growth curve the slope of the straight stretch is the specific growth rate, which sets the doubling time.
rate = (y₂ − y₁) / (x₂ − x₁) μ = (ln N₂ − ln N₁) / (t₂ − t₁) t_d = ln 2 / μ
In words
Rate is rise over run; growth rate is the rise in ln N per hour; doubling time is 0.693 divided by it.
From the growth-curve example above: LB rises from OD600 0.12 at 3 h to 0.5 at 5 h, so μ = (ln 0.5 − ln 0.12) / 2 h = 0.71 per hour and t_d = 0.693 / 0.71 ≈ 0.97 h. The tool draws the points and does not compute slopes or fit curves, so give any rate you report in the figure legend.
Line graph questions, answered
How do I make a line graph from my data?
Type the x values once, then one line per group, or attach the spreadsheet. For example: Time (h): 0, 2, 4, 8. Control: 0.05, 0.12, 0.31, 0.9. Treated: 0.05, 0.09, 0.18, 0.4. Add “mean ± SEM” with replicates written 0.05/0.06/0.05 per point for error bars, “log y axis” for growth data, and axis titles in quotes. The points and ticks are computed from those numbers before anything is drawn.
Line graph or scatter plot: which should I use?
Use a line graph when x is ordered and each point follows from the one before it, as with time, dose or visit; use a scatter plot when each x–y pair is an independent observation. Joining points claims a continuous change between them, which holds for a time course and not for, say, height against weight across people. For paired measurements with a trend line or a correlation, use the scatter plot maker.
Should a line graph show SD or SEM?
SD to describe how much the measurements vary, SEM or a 95% confidence interval to compare means. Whichever you choose, the legend must name it and give n, the number of independent experiments or animals behind each point (Cumming, Fidler and Vaux, J Cell Biol 2007). With n = 3, plotting the individual values is often clearer than any error bar.
When should the y axis of a line graph be logarithmic?
When the values span more than about one order of magnitude and change by multiplication, as in bacterial growth, viral titres or serial dilutions. On a log axis equal ratios take equal distances, so steady exponential growth draws as a straight line. Every value must be above zero: a count of 0 cannot be placed, so the tool keeps a linear axis and says why.
Can a line graph have two y axes?
Not here, and many journals discourage it: where two lines cross on a dual-axis plot depends only on how each axis range was chosen. Ask for two y axes or for separate panels and the tool draws stacked panels on one shared x axis, each with its own y axis and ticks.
Why are my time points not evenly spaced on the x axis?
Because they were not evenly spaced in time. Samples at 0, 1, 2, 4, 8 and 24 h sit at their true positions, so the slope between two points is a real rate. If the x values are visit names rather than times, type them as labels (Baseline, Week 4, Week 12) and they are spaced evenly in the order given.
Should the points be joined by straight lines or a smooth curve?
Straight segments, unless you have a model. A straight segment claims nothing about what happened between measurements; a smooth curve suggests values nobody measured, and a fitted sigmoid needs fitted parameters. Ask for a smooth curve and it passes exactly through every point without overshooting; for a fitted dose-response, fit it in your statistics software and supply the curve's values.
What should the figure legend of a line graph say?
What the points are, what the error bars are, the n behind each point, and what the lines join. For example: “Blood glucose after an oral glucose load in wild-type and db/db mice. Points are means ± SEM, n = 4 mice per group; lines join consecutive time points.” Add the statistical test if the figure marks significance.
How many lines can one line graph hold?
Up to six lines of up to 40 points each here, though four or five is the practical limit for a reader. Each line gets its own colour and marker shape (circle, square, triangle, diamond, inverted triangle, hexagon) and, in the grayscale presets, its own dash pattern. Beyond that, split the conditions into panels.
Is a line chart the same as a line graph?
Yes: line chart is the usual name in spreadsheet and business software, line graph the usual name in school and science. Both plot values against an ordered x axis and join them with lines. A run-sequence plot is a line graph of measurements in the order they were taken, used to spot shifts and drift (NIST/SEMATECH e-Handbook).
Does the AI change my numbers?
No: each point reaches the image model as an exact coordinate. Replicate means, the SD, SEM or 95% CI ends, both axis ranges, the tick labels and the legend order are computed in code from your data on every generation, and the model is told to place each point at its stated value without averaging, smoothing or re-ordering anything.
Sources
- Error bars in experimental biology — Cumming, Fidler and Vaux, Journal of Cell Biology (2007)
- 1.3.3.25. Run-Sequence Plot — National Institute of Standards and Technology (NIST)
Keep exploring
Our other diagram tools, and the guides behind them.
Tools
Make a line graph from your data
Type the time points and values or attach the spreadsheet, and get a journal-style line graph with every point and error bar computed by code.
Make a line graph