Box Plot
Learn to build and use box plots in the Cohort Browser.
An Apollo license is required to use the Cohort Browser. Contact DNAnexus Sales for more information.
How to Build a Box Plot
A box plot can be built to visualize numerical fields. You can recognize numerical fields by the
label in the field list. The same fields can also be visualized using a Histogram.
Supported Field Types
Numerical (Integer)
Numerical (Float)
Using Box Plot in the Cohort Browser
Box plots provides quick statistics on the distribution of a numerical field for the current cohort selection. Hovering over the box plot displays a quick summary of:
- Total number of data points covered by current box plot
- Minimum value
- First quartile
- Median value
- Third quartile
- Maximum value
The box plot does NOT account for or display outliers. All non-null values are used for the display.
In the below example field: Coffee Intake, the box plot covers a total 92,401 patients, among which the maximum value of coffee intake being 42 cups per day.
​
Example Box Plot: Coffee Intake
Some fields have non-numeric values. These special values are counted and displayed separately as a list view. In the below example, there are 89 patients who have value "prefer not to answer" for this field which is 0.09% of the total cohort population (100,000 patients).
View non-numeric values in Box Plots. The total cohort size is 100,000 patients.
Box Plots in Cohort Compare
In compare mode, a color-coded box plot is drawn for each cohort in current session. Hover over each box plot to see summary stats for the cohort represented.
Box plot: "Coffee Intake" in cohort compare mode
Notes on Data Preparation
- Integer
- Integer Sparse
- Float
- Float Sparse
While sparse serial types are supported as fields, the non-numeric values are not included as part of this chart type. Special values in the numerical dimension however will be counted and shown separately as a list view.
Last modified 11d ago
Export as PDF
Copy link