Numeric representation when `options()$scipen` has been changed

[nutterb]

In an e-mail from a user I received the following report

• Trying to produce the latex code for a table from R using pixiedust
  • values in table vary widely from 1.0e4 to 1e-7, therefore for consistency I've forced them all to scientific notation using "options(scipen=-10)"
  • however when I run this code at the end

stuff <- results.print %>% dust %>% sprinkle_print_method("latex") %>% sprinkle(font_size=6) %>% print(asis = FALSE) %>% cat()

• the latex code created has the fonts set as "6e+00pt" which latex doesn't like
• alternatively if I reset the scipen back to the default by running "options(scipen=0)" , just prior to this code then the font specification in the latex code comes out correctly, but the values in the table are changed back to a mix of scientific and regular notation

results.print <- results2[289:325, ] head(results.print) timestep walk-area date rain temp Ep Qt 289 289 BC 2013-10-16 2.85e-02 9.25e+00 8.58e-04 4.500e+02 290 290 BC 2013-10-17 1.05e-03 1.10e+01 9.44e-04 7.519e+03 291 291 BC 2013-10-18 3.33e-02 1.09e+01 9.23e-04 4.736e+04 292 292 BC 2013-10-19 2.38e-02 1.29e+01 1.03e-03 1.273e+06 293 293 BC 2013-10-20 1.59e-02 1.20e+01 9.58e-04 3.581e+02 294 294 BC 2013-10-21 1.87e-02 1.24e+01 9.74e-04 4.773e+03 qt qo qs qv Smean 289 2.583e-05 2.115e-05 4.682e-06 2.714e-02 3.606e-02 290 4.317e-04 6.860e-07 4.310e-04 1.934e-04 3.630e-02 291 2.719e-03 2.305e-03 4.143e-04 3.009e-02 6.618e-03 292 7.307e-02 1.479e-02 5.827e-02 8.069e-03 5.682e-02 293 2.056e-05 7.015e-06 1.354e-05 1.482e-02 4.201e-02 294 2.740e-04 1.143e-04 1.598e-04 1.767e-02 2.451e-02 > stuff <- results.print %>%

• dust %>%
• sprinkle_print_method("latex") %>%
• sprinkle(font_size=6L) %>%
• print(asis = FALSE) %>%
• cat() \setlength{\tabcolsep}{6e+00pt} \begin{table} \centering

\label{tab:pixie-0e+00}\begin{tabular}{lllrrrrrrrrr} \multicolumn{1e+00}{l}{timestep} & \multicolumn{1e+00}{l}{walk-area} & \multicolumn{1e+00}{l}{date} & \multicolumn{1e+00}{r}{rain} & \multicolumn{1e+00}{r}{temp} & \multicolumn{1e+00}{r}{Ep} & \multicolumn{1e+00}{r}{Qt} & \multicolumn{1e+00}{r}{qt} & \multicolumn{1e+00}{r}{qo} & \multicolumn{1e+00}{r}{qs} & \multicolumn{1e+00}{r}{qv} & \multicolumn{1e+00}{r}{Smean} \

The following elements need to be guarded against scientific notation representations

• [x] font size
• [x] border thickness
• [x] cell height
• [x] cell width
• [x] tabcolsep
• [x] column span
• [x] row span
• [x] label numbering
• [x] longtable (when a numeric is passed, it should be an integer)
• [x] merge_rowval (should be an integer)
• [x] merge_colval (should be an integer)
• [x] pad
• [x] rotate_degree

Probably the easiest thing to do is store them in the internal structure as integers? Does that work for values that could theoretically accept a decimal?

[nutterb]

Under 0.7.4,

library(pixiedust)
DF <- 
structure(list(timestep = 289:294, walk.area = c("BC", "BC", 
"BC", "BC", "BC", "BC"), date = c("2013-10-16", "2013-10-17", 
"2013-10-18", "2013-10-19", "2013-10-20", "2013-10-21"), rain = c(2.85e-02, 
1.05e-03, 3.33e-02, 2.38e-02, 1.59e-02, 1.87e-02), temp = c(9.25e+00, 
1.1e+01, 1.09e+01, 1.29e+01, 1.2e+01, 1.24e+01), Ep = c(8.58e-04, 
9.44e-04, 9.23e-04, 1.03e-03, 9.58e-04, 9.74e-04), Qt = c(4.5e+02, 
7.519e+03, 4.736e+04, 1.273e+06, 3.581e+02, 4.773e+03), qt = c(2.583e-05, 
4.317e-04, 2.719e-03, 7.307e-02, 2.056e-05, 2.74e-04), qo = c(2.115e-05, 
6.86e-07, 2.305e-03, 1.479e-02, 7.015e-06, 1.143e-04), qs = c(4.682e-06, 
4.31e-04, 4.143e-04, 5.827e-02, 1.354e-05, 1.598e-04), qv = c(2.714e-02, 
1.934e-04, 3.009e-02, 8.069e-03, 1.482e-02, 1.767e-02), Smean = c(3.606e-02, 
3.63e-02, 6.618e-03, 5.682e-02, 4.201e-02, 2.451e-02)), .Names = c("timestep", 
"walk.area", "date", "rain", "temp", "Ep", "Qt", "qt", "qo", 
"qs", "qv", "Smean"), class = "data.frame", row.names = c(NA, 
-6L))

options(scipen = -10)

dust(DF) %>%
  sprinkle_print_method("html") %>%
  sprinkle(font_size=6) %>%
  sprinkle(cols = 1:2,
           rows = 1:2,
           merge_rowval = 2,
           merge_colval = 2,
           merge = TRUE)