NA/NAN Argument Error in R loop

Question

I am new in R and I would like to get some help. I have an R code with user defined function called Plot_linear_fit below that I ran without any errors.

 ReportDateRange <- c("2010-11-01", "2017-01-31")
 trendDateRange1 <- c("2015-01-01", "2015-08-31")
 trendDateRange2 <- c("2015-01-01", "2016-10-31")
 trendDateRange3 <- c("2015-01-01", "2016-06-30")
 plotDate1 <- c("2011-01-01")
 plotDate2 <- c("2015-01-01")
 plotDate3 <- c("2013-01-01")
 numoftrends <- 3

 TRx.Plot1 <- Plot_linear_fit(Sum.TRx,
                              trendDateRange1,
                              plotDate1, 
                              ReportDateRange)

 TRx.Plot2 <- Plot_linear_fit(Sum.TRx,
                              trendDateRange2,
                              plotDate2, 
                              ReportDateRange)

 TRx.Plot3 <- Plot_linear_fit(Sum.TRx,
                              trendDateRange3,
                              plotDate3, 
                              ReportDateRange)

I would like to put these codes in a loop using lapply however when I try to run it, I am getting the NA/NAN argument

 ReportDateRange <- c("2010-11-01", "2017-01-31")
 trendDateRange1 <- c("2015-01-01", "2015-08-31")
 trendDateRange2 <- c("2015-01-01", "2016-10-31")
 trendDateRange3 <- c("2015-01-01", "2016-06-30")
 plotDate1 <- c("2011-01-01")
 plotDate2 <- c("2015-01-01")
 plotDate3 <- c("2013-01-01")
 numoftrends <- 3

 lapply(1:numoftrends, function(j) {
    paste0('TRx.Plot', j) <- Plot_linear_fit(Sum.TRx,
                                 paste0("trendDateRange", j),
                                 paste0("plotDate", j),
                                 ReportDateRange)

     })

I am not so sure what is wrong when you put this in lapply. The output of the function Plot_linear_fit is a dataframe. Thank you for your help.

You have to distinguish between an unquoted name as a literal character string, and the same name as a quoted character string. Unfortunately, R makes it even more confusing by sometimes being very tolerant to arguments being passed either as literal names or as character strings, some other times not at all — Aurèle, Apr 27 '17 at 16:43
Thank you for your time. I do apologise, I am new to R and am still learning. How would you code this? The TRx.Plot1, TRx.Plot2 and TRx.Plot3 are all data frames. I want to use data frame in ggplot. — aotearoa, Apr 27 '17 at 16:49
I'm fairly sur there is no plot_linear_fit function in base R. — IRTFM, Apr 27 '17 at 17:04
Thank your time. The plot_linear_fit is a user defined function. — aotearoa, Apr 27 '17 at 17:05

score 2 · Accepted Answer · edited May 23 '17 at 12:02

The minimal fix would be (I guess, untested code because you didn't provide a MRE):

lapply(1:numoftrends, function(j) {
  assign(paste0('TRx.Plot', j),
         Plot_linear_fit(Sum.TRx,
                         get(paste0("trendDateRange", j)),
                         get(paste0("plotDate", j)),
                         ReportDateRange),
         pos = .GlobalEnv)
})

But it would be more idiomatic to drop assignments in the global workspace altogether, and work directly with lists:

list_of_plots <- lapply(1:numoftrends, function(j) {
  Plot_linear_fit(Sum.TRx,
                  get(paste0("trendDateRange", j)),
                  get(paste0("plotDate", j)),
                  ReportDateRange)
})

Thanks for your time. This is what I am looking for. Have a great day! — aotearoa, Apr 28 '17 at 13:02

NA/NAN Argument Error in R loop

1 Answers1