Suggest that we add a variable description file. This could simply be table1Long.csv (not sure that is the right name) and include a range of descriptive statistics that are needed, see below, or just reference to inform how the algorithm was created.
- We may want to specify important descriptive data that is included, such as max, min, median, min.
- This file would be used as:
- a reference file.
- for model visualization. The current model visualization tool uses percentiles of exposures or all continuous predictors by sex and age (separately).
- possibily for centering values and calibration. However, it makes sense to keep the centered values in the center.csv file.
- There may be more than one descriptive file or it could be large. For example, there could be a descriptive stats file for derivation and validation data.
Suggest that we add a variable description file. This could simply be table1Long.csv (not sure that is the right name) and include a range of descriptive statistics that are needed, see below, or just reference to inform how the algorithm was created.