I am trying to sum probes count within subgroup of a dataframe in R -

this question has answer here:

• idiomatic r code partitioning vector index , performing operation on partition 3 answers

i biochemistry student working breast cancer data (copy number probes). have following data frames primary breast tumors:

patient   chrom   start    end       probecount 1         1       51599    62640     8 1         1       88466    16022503  8676 1         2       2785     285255    186 1         2       290880   4178544   2903 ... 2         1       51599    4098530   1282 2         1       4101675  46753618  25229 2         2       2785     36178040  25931 2         2       36185342 36192717  21 ... 

i add add fifth column add total of probecounts each chrom of each patient:

patient   chrom   start    end       probecount   total 1         1       51599    62640     8            8684 1         1       88466    16022503  8676         8684 1         2       2785     285255    186          3089 1         2       290880   4178544   2903         3089 ... 2         1       51599    4098530   1282         26511 2         1       4101675  46753618  25229        26511 2         2       2785     36178040  25931        25952 2         2       36185342 36192717  21           25952 ... 

there must simple function this. aggregate? appreciate if please give me hint. thank you!

one way dplyr package following. data frame called mydf.

library(dplyr) group_by(mydf, patient, chrom) %>% mutate(whatever = sum(probecount))  #source: local data frame [8 x 6] #groups: patient, chrom # #  patient chrom    start      end probecount whatever #1       1     1    51599    62640          8     8684 #2       1     1    88466 16022503       8676     8684 #3       1     2     2785   285255        186     3089 #4       1     2   290880  4178544       2903     3089 #5       2     1    51599  4098530       1282    26511 #6       2     1  4101675 46753618      25229    26511 #7       2     2     2785 36178040      25931    25952 #8       2     2 36185342 36192717         21    25952 

if data large, may want use data.table.

library(data.table) setdt(mydf)[, whatever := sum(probecount), = list(patient, chrom)][]  #   patient chrom    start      end probecount whatever #1:       1     1    51599    62640          8     8684 #2:       1     1    88466 16022503       8676     8684 #3:       1     2     2785   285255        186     3089 #4:       1     2   290880  4178544       2903     3089 #5:       2     1    51599  4098530       1282    26511 #6:       2     1  4101675 46753618      25229    26511 #7:       2     2     2785 36178040      25931    25952 #8:       2     2 36185342 36192717         21    25952