file.conf
to the end of the file.name
, right before the disease status (last) column. The output will contain only the patients for which confounding variables exist (other patients will be omitted), so new family file will be written. This function is similar to pre7.add.conf.var
, only it avoids having to load up into memory the file.name
(since this file can be very large).
pre7.add.conf.var.unix(file.name, dir.file, file.fam, dir.fam = dir.file,
file.conf, dir.conf = dir.file, file.out, fam.out = file.fam, dir.out)
file.name
can be found.
file.fam
can be found.
patientID1 1 2 ... patientID2 3 1 ... patientID3 2 2 ... ...- Column 1: patient ID, exactly the same names should appear in file.fam; * order does not matter; * some patients may be missing; * no new patients should appear in file.conf (if they don't exist in file.fam) - Column 2: the confounding variable must have no more than 3 different values. - Other columns are optional, may be included if there are more confounding variables (3 categories each) - No header - Tab separated - No missings or NAs
file.conf
can be found.
file.name
, plus confounding variables, only for the patients mentioned in file.conf
.
file.out
and fam.out
should be saved.
pre7.add.conf.var
, pre6.merge.genos
,
pre8.split.train.test
, pre8.split.train.test.batch
print("See the demo 'gendemo'.")
Run the code above in your browser using DataLab