handle column names that aren't valid?

nwis_data <- readNWISuv(siteNumbers = "01389005", 
        parameterCd = "00300", startDate = "2006-09-30T23:00Z", 
        endDate = "2017-01-01T00:00Z")
names(nwis_data)
 [1] "agency_cd"                            "site_no"                             
 [3] "dateTime"                             "X_,from.right.intake_00300_00000"    
 [5] "X_,from.right.intake_00300_00000_cd"  "X_,from.middle.intake_00300_00000"   
 [7] "X_,from.middle.intake_00300_00000_cd" "X_,from.left.intake_00300_00000"     
 [9] "X_,from.left.intake_00300_00000_cd"   "tz_cd" 

and the , breaks a lot of parsers in R, for example dplyr::select

perhaps a make.names call on the headers?

make.names("X_,from.left.intake_00300_00000")
[1] "X_.from.left.intake_00300_00000"