Remove elements that occur before the first bin when splitting by bins

This change is in-line with the current treatment of the elements that
occur after the last bin.

Further, this fixes a minor "bug" where bin indecees are used to
determine bin labels by indexing into a list of labels. Elements before
the first bin received bin index 0, however, R is 1-indexed. Note: This
"bug" did not cause problems because the bin with the incorrect label
will be ignored later anyways.

Works towards se-sic#267.

Signed-off-by: Maximilian Löffler <s8maloef@stud.uni-saarland.de>

util-split.R

@@ -1016,6 +1016,8 @@ split.data.by.time.or.bins = function(project.data, splitting.length, bins, spli
                         findInterval(df[["date"]], bins.date, all.inside = FALSE)
             ## split data according to df.bins
             df.split = split(df, df.bins)
+            ## remove data before the first bin
+            df.split = df.split[names(df.split) != "0"]
             ## add proper labels/names
             names(df.split) = sapply(as.integer(names(df.split)), function(bin) bins[bin])
             return(df.split)