Scatter plot and Bubble chart
In Anaplan, a Scatter plot is a type of plot which displays values for two sets of data series, along two axes. The distribution of the dots across the axes demonstrate the existence (or absence) of correlation between the events. Bubble charts exist as an added dimension of Scatter plots, replacing the plotted dots of the Scatter plot with 'bubbles' representing a third series of data, typically meaning size or volume of a dimension related to the first two.
Scatter plots display a series of points across horizontal and vertical axes, based on how two data series influence each other. Scatter plots are useful because they can show the extent of correlation, if any, between the observed variables. If no correlation exists between them, the points appear randomly scattered on the coordinate plane. If a large correlation exists, the points concentrate near a straight line across the axes.
Bubble charts add another dimension to scatter plots by including a third, related, series of data, which demonstrates size. This emphasizes the size of the dimension by turning the dots (plots on the chart) into bubbles. In Anaplan, the size of the bubble usually corresponds to the market size.
In both cases, a single series of data is charted from the first two columns of line items (three in the case of a bubble chart) in your grid. You may need to hide columns to establish the required adjacency.
Switch on Trendlines to more easily draw conclusions. Then, you can use:
- Linear trendlines, for simple linear data sets and produces a best-fit straight line. This is used to demonstrate constant progression or decline
- Polynomial trendlines, when values fluctuate and creates an up-and-down curved line. This is used to analyze gains and losses in a large data set
- Exponential trendlines, with larger datasets with continuously fluctuating values. Due to their nature, Exponential trendlines cannot contain nonpositive (greater-than-zero) values in the Y-axis and will be unavailable if these values are present
- Logarithmic trendlines, to facilitate the display of sudden increases or decreases in the trend, but tend to level out. Due to their nature, Logarithmic trendlines cannot contain nonpositive (greater-than-zero) values in the Z-axis and will be unavailable if these values are present
What can I use scatter plots and bubble charts for?
The scatter plot and the bubble chart are best used when comparing paired numerical data, not necessarily associated to a time dimension. The plot of markers across the visualization will allow you to extrapolate the displayed data and derive conclusions based on your use case. If the markers are grouped tightly together, then there's a high correlation. If the markers on the chart are apparently random in distribution, then there's low to no correlation in the data. You can do the following with this type of chart:
- Determine the root cause of a given scenario, by comparing numerical data
- Visualize the relationship between your two data series
- Identify where or who are the outliers in the data set
- Determine a trend in the correlation of data
- Identify opportunities and/or risk vs. return, in a competitive scenario
Apart from generic chart configuration options, as described in Add a chart card to a board, scatter plots and bubble charts have unique configuration options and requirements. These include:
- Define minimum and maximum values for the chart axes
- Create a bubble chart via the Card ConfigurationChart tab, by toggling right on Size
A scatter plot or bubble chart may not be the best option when:
- you're handling a small amount of data
- you're handling non-numerical data