Annotations
Use the Annotate left side panel to add annotations to your chart. Annotations can help you highlight trends and patterns in your data visualizations.
Trend Lines
To add a trend line to your chart, open the Annotate left side panel, then click Add line under Trend Lines.
A trend line can be customized in the following ways:
- Type Choose the type of line to draw on top of your chart. Choices include Average, Custom Value and Linear Regression among others.
- Value For the custom value trend line. Enter in a value to highlight on the Y-Axis
- Label The text to label a line with (not available on all line types)
- Text Color Choose to display either light or dark text in the label
- Stroke Choose the line stroke. Options include solid, small and large dotted
- Thickness Choose the line thickness
- Color Choose the color of the trend line and its label (if it has one)
The different trend line types are described below.
Custom Value
Use a custom value trend line to showcase a constant value on the Y-Axis.
Use cases
- Highlight an important custom metric
Average
The average (mean) line shows the arithmetic mean of your data points, providing a quick visual reference for the central tendency of your dataset.
Use cases
- Identifying whether data points are above or below average
- Comparing different segments of data against the overall average
- Spotting outliers or unusual patterns
Median
The median line represents the middle value of your dataset when arranged in order, effectively dividing your data into two equal halves.
Use cases
- Visualizing the center of your data distribution
- Identifying skewness in your dataset
- Providing a robust measure of central tendency less affected by outliers
Linear Regression
A linear regression line shows the best straight-line fit through your data points, indicating overall linear trends.
Use cases
- Identifying and visualizing linear relationships between variables
- Predicting values based on the linear trend
- Comparing actual values against the predicted linear relationship
LOESS Regression
LOESS (Locally Estimated Scatterplot Smoothing) regression creates a smooth line through your data points, showing local trends and patterns.
Use cases
- Visualizing non-linear relationships in your data
- Identifying complex patterns that might be missed with linear regression
- Smoothing out noise in your dataset to focus on overall trends
Choosing the best trend line for your data
- Use average/median lines for reference points.
- Use LOESS for complex, non-linear patterns.
- Use linear regression for straight-line trends.