unable to perform calculations using r data.table package

Question

I have a huge data frame, last 30 rows are below:

libary(data.table)

dput(p)

structure(list(DATE = structure(c(1367516015, 1367516045, 1367516075, 
1367516105, 1367516135, 1367516165, 1367516195, 1367516225, 1367516255, 
1367516285, 1367516315, 1367516345, 1367516375, 1367516405, 1367516435, 
1367516465, 1367516495, 1367516525, 1367516555, 1367516585, 1367516615, 
1367516645, 1367516675, 1367516705, 1367516735, 1367516765, 1367516795, 
1367516825, 1367516855, 1367516885), class = c("POSIXct", "POSIXt"
), tzone = ""), LPAR = structure(c(6L, 6L, 6L, 6L, 6L, 6L, 6L, 
6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 
6L, 6L, 6L, 6L, 6L, 6L, 6L), .Label = c("cigp01a4a004", "cigp01b4a002", 
"cigp01b4a004", "cigp04a4a002", "cigp04a4a004", "cigp04b4a002", 
"cigp04b4a004"), class = "factor"), ENT = c(0.5, 0.5, 0.5, 0.5, 
0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 
0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5
), USR_SYS_CPU_PCT = c(79L, 80L, 77L, 77L, 77L, 76L, 79L, 82L, 
81L, 80L, 79L, 77L, 77L, 77L, 79L, 79L, 80L, 82L, 82L, 83L, 80L, 
81L, 80L, 78L, 78L, 83L, 86L, 87L, 88L, 87L), ENT_PCT = c(706.8, 
693.8, 570.1, 641.5, 558.5, 601.5, 674.3, 742.3, 668.9, 722.6, 
679.1, 677.2, 548.5, 644.6, 689.3, 716.1, 709.5, 767.3, 753.7, 
786.4, 684.2, 735.1, 688.2, 676.6, 645.6, 788, 859.5, 832.6, 
883.1, 872.2), PHYSICAL_CPU_USED = c(3.53, 3.47, 2.85, 3.21, 
2.79, 3.01, 3.37, 3.71, 3.34, 3.61, 3.4, 3.39, 2.74, 3.22, 3.45, 
3.58, 3.55, 3.84, 3.77, 3.93, 3.42, 3.68, 3.44, 3.38, 3.23, 3.94, 
4.3, 4.16, 4.42, 4.36), PROC_QUE = c(12L, 13L, 19L, 16L, 11L, 
13L, 17L, 14L, 9L, 10L, 12L, 13L, 16L, 14L, 22L, 17L, 17L, 17L, 
26L, 26L, 15L, 43L, 9L, 11L, 12L, 7L, 31L, 26L, 27L, 23L), RELATIVE_CORES = c(3.53, 
3.47, 2.85, 3.21, 2.79, 3.01, 3.37, 3.71, 3.34, 3.61, 3.4, 3.39, 
2.74, 3.22, 3.45, 3.58, 3.55, 3.84, 3.77, 3.93, 3.42, 3.68, 3.44, 
3.38, 3.23, 3.94, 4.3, 4.16, 4.42, 4.36), USED_CORES = c(2.7887, 
2.776, 2.1945, 2.4717, 2.1483, 2.2876, 2.6623, 3.0422, 2.7054, 
2.888, 2.686, 2.6103, 2.1098, 2.4794, 2.7255, 2.8282, 2.84, 3.1488, 
3.0914, 3.2619, 2.736, 2.9808, 2.752, 2.6364, 2.5194, 3.2702, 
3.698, 3.6192, 3.8896, 3.7932)), .Names = c("DATE", "LPAR", "ENT", 
"USR_SYS_CPU_PCT", "ENT_PCT", "PHYSICAL_CPU_USED", "PROC_QUE", 
"RELATIVE_CORES", "USED_CORES"), class = "data.frame", row.names = c(NA, 
-30L))

when I like calcualate some values using data.table as below:

p<-data.table(p)
p<-p[,RELATIVE_PERCENT:=ifelse(ENT_PCT>100, (USED_CORES/ENT)*100, USR_SYS_CPU_PCT), by= c("DATE", "LPAR")]

I get this error:

Error in `[.data.table`(x, , `:=`(RELATIVE_PERCENT, ifelse(ENT_PCT > 100,  : 
  Type of RHS ('integer') must match LHS ('double'). To check and coerce would
  impact performance too much for the fastest cases. Either change the type of
  the target column, or coerce the RHS of := yourself (e.g. by using 1L instead
  of 1)

what does this error mean? How can I get around this error?

Answer 1

The problem is that your ifelse statement returns integer type for some values and numeric (double) for some other entries. And data.table complains about the mismatch in the column type as it expects the coercion to be performed by the user (for performance reasons as given in the error). So, just wrap it around with as.numeric so that all values will be converted to double.

p <- p[,RELATIVE_PERCENT := as.numeric(ifelse(ENT_PCT>100, (USED_CORES/ENT)*100, 
                      USR_SYS_CPU_PCT)), by= c("DATE", "LPAR")]