Comparison analysis refers to data differences caused by other influencing factors (comparison conditions) under the same data standards (identical dimension metrics). The purpose of comparison analysis is to further investigate the reasons behind these differences after identifying them, in order to find ways to optimize. DataWind provides a comparison analysis component that supports comparing two or more groups of data under the same dimensions.
Example

2.1 Common usage scenarios #

Simultaneously compare the prices and price differences across various categories for different regions in a single chart.

2.2 Usage process example #

Configure a table with "type" as the dimension and "price" as the metric.

Step 1: Click the "Analysis" tab in the middle panel and select "Comparison".

Step 2: Drag the dimension field to be used as the comparison condition and configure it.

Step 3: After specifying the calculation type, click "Apply".
To display only the prices for "Shanghai" and "Beijing", simply select "Original Value".

This completes a basic comparison analysis process. To adjust the style, modify the display name and data format on the capsule (the UI element displaying the field), or change the chart style to another chart type.

3.1 Add comparison conditions #

Multiple dimensions can be added to a comparison condition, and they have an "and" relationship.
By default, two comparison condition configurations are provided. To add more comparison conditions, click the add button.

By default, "Comparison Condition 1" serves as the baseline condition, indicated by an icon next to the title.

3.2 Copy comparison conditions #

When comparison conditions are similar, use the copy function to quickly add comparison conditions. When the mouse hovers over the panel, a "Copy" button appears in the upper right corner of the title. Click it to directly copy the current comparison condition.

3.3 Set active metrics #

By default, all metrics in the metrics panel are set as active metrics. To modify the selection, click the dropdown list to reselect the active metrics.

3.4 Switch baseline condition #

By default, "Comparison Condition 1" is set as the baseline condition. To change the baseline condition, click the dropdown list to switch the baseline condition. After switching, the "Baseline" label will appear next to the title of the corresponding comparison condition.

3.5 Select calculation type #

We provide a total of four calculation types:

• Original value: The metric value under the comparison condition
• Difference value: Comparison condition value minus baseline condition value
• Difference percentage: (Comparison condition value minus baseline condition value) / baseline condition value * 100%
• Ratio: Comparison condition value divided by baseline condition value
• Confidence interval: An interval defined by the upper and lower confidence limits of the statistic

The trigger conditions for the confidence interval are relatively strict:

• Only confidence interval calculations for two comparison conditions are supported
• A field for calculating the confidence interval must be specified, and this field must be a numeric dimension
• When entering the field, it must be an expression, not a field name

3.6 Original field capsules and derived field capsules #

After applying the comparison condition, the active metrics will split into a series of derived metrics as required.

Unlike original metrics, derived metric capsules have a lighter color and a dashed border. They cannot be deleted individually. To delete derived metrics, deselect them in the configuration panel on the left.
At this point, the original metric capsules serve only as placeholders and are not displayed in the chart. (Because they have already been split into the original values for comparison)
Each derived metric supports actions such as modifying the display name and changing the data format.
Capsules can be reordered by dragging and dropping.