Changes to parse the columns in input tree data.frame. The behaviour has been simplified to accept length measurements. There used to be logic to use SNVs and PGA as branch lengths, but this was removed to simplify the input. This is consistent with the design decision to focus on plotting the data, and exclude any logic to analyze the input data.
Now, all columns that contain "length" are used as branch lengths. For example length1 and 2length will both be used. The plan is to use all matching columns as branches, but the plotting function does not currently support this. The prep functions use the first 2 matches, and display a message if more columns are passed. This allows us to establish the ideal API, while adding the functionality later.
Any length columns that contain NA values are not used (after displaying a warning message).
If no branch length columns are passed, a default value of 1 is used for all. This seems to be the most reasonable length based on the existing example plots.
Changes to parse the columns in input tree
data.frame. The behaviour has been simplified to accept length measurements. There used to be logic to use SNVs and PGA as branch lengths, but this was removed to simplify the input. This is consistent with the design decision to focus on plotting the data, and exclude any logic to analyze the input data.Now, all columns that contain "length" are used as branch lengths. For example
length1and2lengthwill both be used. The plan is to use all matching columns as branches, but the plotting function does not currently support this. The prep functions use the first 2 matches, and display a message if more columns are passed. This allows us to establish the ideal API, while adding the functionality later.Any
lengthcolumns that containNAvalues are not used (after displaying a warning message).If no branch length columns are passed, a default value of
1is used for all. This seems to be the most reasonable length based on the existing example plots.