3

I have a large data set in Excel (50k rows, 138 columns). 80 of the columns are date/time values which I want to use as POSIXct type in R.

Option 1: Use XLConnect or xlsx with the colTypes parameter. Fails due to size of the file with heap size errors.

Option 2: Save as csv and read file, but Excel drops the time values when save as CSV.

Option 3: read into R using openxlsx package with the detectDates option. This loses the time values.

Option 4: read into R using openxlsx package with detectDates=false. This returns the date/time values as numeric.

This last option seems to be the most viable, but how do I then conveniently convert 80 columns to POSIXct? I have a vector of classes that I want for each of the columns that would be provided to the colTypes parameter, but now want to apply this in post-processing after reading the data frame.

ZygD
  • 22,092
  • 39
  • 79
  • 102
sch56
  • 361
  • 1
  • 11

0 Answers0