Saving analyses to file v5

The following instructions work in the DataSHIELD test environment only.

Write a table to .csv

  • use the write.csv or write.table R functions to write tables to file
# output to the R console
> ds.table2D(x='D$DIS_DIAB', y='D$GENDER')
# prints to the current working directory into a file called test_table.csv
> write.csv(x=ds.table2D(x='D$DIS_DIAB', y='D$GENDER'), file = "test_table.csv") 

Write an image to .png, .pdf, .ps

In R it is possible to print sequential images on sequential pages.

  • To do this an image writing function such as png(), pdf() or postscript() needs to be opened.
  • Each file to be written must be put on seperate lines
  • The print device needs closing.
  • All images are written to the current working directory in R.
# a png example
png(filename = "image_name.png", width = 900, height = 700, units = "px", pointsize = 12,  bg = "white", type = "cairo")
ds.histogram(x='D$LAB_HDL')
dev.off()
  • The pdf example prints the output of two seperate ds.histogram commands onto seperate pages. Each line is output to a seperate page in the pdf.
# a pdf example
pdf(file = "image_name.pdf")
ds.histogram(x='D$LAB_HDL')
ds.histogram(x='D$LAB_HDL', type='split')
dev.off()
# a postscript example
postscript(file = "image_name.ps", paper = "special", width = 9, height = 7, onefile = TRUE, horizontal = FALSE)
ds.histogram(x='D$LAB_HDL')
ds.histogram(x='D$LAB_HDL', type='split')
dev.off()

Write the R console screen to file

In some cases the non-disclosive data used to create an image may be required, this can be printed to file using the capture.output R function:

# print to file the non-disclosive data that comprises a histogram
out <- capture.output(ds.histogram(x='D$LAB_HDL'))
cat(out, file="name.txt", sep="\n", append=TRUE)

All arguments to customise the above functions can be found in their respective function manual pages.

DataSHIELD Wiki by DataSHIELD is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. Based on a work at http://www.datashield.ac.uk/wiki