A radar chart compares multiple objects across shared metrics by placing each metric on a radial axis. It is useful for model comparison, sample scoring, multi-index evaluation, and performance summaries.
Prepare a directly mappable table
Radar charts need several comparable numeric metrics for each object. The metrics should use a consistent scale or be normalized before plotting.
In wide tables, each row can be one object and each metric can be one numeric column. In column-oriented tables, each object can be a column. In long tables, prepare object, metric, and value fields.
Practice data · illustrates column structure
method
accuracy
precision
recall
Model A
0.91
0.88
0.86
Model B
0.87
0.92
0.84
Model C
0.89
0.85
0.93
Create a radar chart
Use a radar chart when you need to compare several methods, samples, or groups across the same set of metrics. Choose the data layout first, then map the object labels and radar axes.
Step 1
Choose range and output
Confirm the data range, then choose a new figure or add the plot to an existing figure.
