Choose bar, line, or point marks according to the comparison or relationship people need to understand.
Choose a familiar chart form when it can communicate the data clearly. The mark type should explain a relationship, not merely provide visual variety. The examples here describe chart design; they are not Liquidify components or API examples.
| Question | Useful starting point | Check before using it |
|---|---|---|
| How do category totals compare? | Bars | Categories use a shared scale and clearly labeled units. |
| How does a measure change over time? | A line | Observation order and the meaning of connections are clear. |
| How do two measurements relate? | Points | Both axes identify the measured variables and units. |
For example, bars can compare the number of requests received on each weekday. A line can show changes in a queue measured at regular intervals. Points can compare request size with handling time, keeping each observation distinct. These are different questions and should not share a title merely because they use related records.
Figure in preparation
Choose marks from the question
Compare category totals, change over time, and paired observations using different marks.
Use a common baseline for bars when their lengths represent amounts. A zero baseline normally preserves that comparison. A line chart can use a narrower range when small changes matter, provided the range is explicit. Review Axes and labels before changing a scale.
A connection between points suggests continuity. Decide how missing observations are represented before drawing a line through them; missing data is not automatically zero. Explain aggregation, such as daily totals or weekly averages, so that readers know what each mark represents.
Combine marks only when the combination adds information. Adding points to a line can identify the actual observations, but a second encoding should not obscure the trend or imply measurements that were never collected.
Avoid packing every series and label into the initial view. Let people choose a relevant period or subset when the dataset is large. Keep the current selection visible and preserve the meaning of series colors and labels across views.
Provide a summary and relevant exact values outside interactions that depend on a pointer. If individual marks can be selected, review their target size and keyboard navigation in the chosen renderer. Do not assume that every mark must become a separate Tab stop.
Liquidify has no bar, line, or scatter chart API. A separate chart implementation must provide marks, scales, and accessible interaction. Use Gauge only for one bounded measurement; it does not replace a series chart. The overview distinguishes charting from other data displays.