Scatter Plot
A scatter plot places each record as a dot, using one number for the horizontal position and another for the vertical position. It is the classic way to see whether two variables are related, spot clusters, and find outliers.
The makechart.io scatter chart maker also lets you color dots by category, size them by a third value and show a regression line. Paste or upload your data, bind the columns, and download the data visualization as SVG or PNG.
Data format
One row per data point, with at least two numeric columns. Category, size and name columns are optional and add more information to each dot.
| Field | Type |
|---|---|
X coordinaterequired | Number Horizontal position of each dot. |
Y coordinaterequired | Number Vertical position of each dot. |
categoryoptional | Category Groups dots by color and adds a legend. |
sizeoptional | Number Dot size, turning the plot into a bubble-style scatter. |
nameoptional | Text or category Name of each point, shown in the tooltip. |
labeloptional | Text Text shown as the dot label. The name is used if omitted. |
Orderoptional | Order Sequence of the points, used when connecting them with a trend line. |
| name | height (cm) | weight (kg) | group |
|---|---|---|---|
| Amy | 158 | 49 | A |
| Ben | 175 | 70 | B |
| Cora | 163 | 55 | A |
| Dan | 181 | 78 | B |
| Eve | 169 | 61 | A |
When to use
- Checking whether two numeric variables are correlated, such as advertising spend and sales.
- Spotting clusters and outliers in hundreds of data points.
- Comparing groups by coloring dots by category.
- Showing a third variable by dot size.
When not to use
- Only one numeric variable; use a bar chart or histogram.
- Categories on both axes; a heat map style table or grouped chart reads better.
- Telling a story about change over time for a single series; a line chart is clearer.
Tips
- A regression line is shown by default; turn regression off in the Style panel if you do not need it, or turn on trend to connect points by Order.
- Set the scale type of xaxis or yaxis to log scale when values span several orders of magnitude.