no dataset id with as.data.frame

[fkgruber]

Hi, I tried the multiple models multiple datasets example with the as.data.frame function but it did not return the dataset id. I imagine that given that this example have multiple datasets that I would want to distinguish between datasets.

samps <- create_sim_samples(10, 100, 100, "all") mdat <- mmdata(samps[["scores"]], samps[["labels"]], modnames = samps[["modnames"]], dsids = samps[["dsids"]])

 ## Generate an mscurve object that contains ROC and Precision-Recall curves

mmcurves <- evalmod(mdat, raw_curves = TRUE)

as.data.frame(mmcurves)

thanks Fred

[takayasaito]

Thanks for your suggestion. Precrec offers two formats for as.data.frame.

1. x, y, modname, dsid, type
2. x, y, ymin, ymax, modname, type

The as.data.frame function currently shows the format 2 for the mmcurves object by default since it ignores the raw_curves option that has been set in the evalmod function. I'll fix this issue and include it in the next release.

For the time being, you can explicitly set the raw_curves option with as.data.frame.

> # Create sample datasets with 100 positives and 100 negatives
> samps <- create_sim_samples(10, 100, 100, "all")
> mdat <- mmdata(samps[["scores"]], samps[["labels"]],
+                modnames = samps[["modnames"]],
+                dsids = samps[["dsids"]])
> 
> # Generate an mscurve object that contains ROC and Precision-Recall curves
> mmcurves <- evalmod(mdat, raw_curves = TRUE)
> 
> # Convert mmcurves to a data frame **Set raw_curves = TRUE again here!!!**
> mmcurves.df <- as.data.frame(mmcurves, raw_curves = TRUE)
> 
> # Show data frame
> head(mmcurves.df)
      x y modname dsid type
1 0.000 0  random    1  ROC
2 0.001 0  random    1  ROC
3 0.002 0  random    1  ROC
4 0.003 0  random    1  ROC
5 0.004 0  random    1  ROC
6 0.005 0  random    1  ROC

[fkgruber]

thanks that works