error in LoadLidar()

[ptompalski]

When loading a pointcloud with LoadLidar() or readLas() following error occurs:

Error in as.data.table.matrix(xi, keep.rownames = keep.rownames) : 
  4 arguments passed to .Internal(nchar) which requires 3

error does not occur when
return(data.table(mm, Intensity, ReturnNumber,NumberOfReturns,Classification,ScanAngle,R,G,B,gpstime))
is changed into
return(data.frame(mm, Intensity, ReturnNumber,NumberOfReturns,Classification,ScanAngle,R,G,B,gpstime))
readLas.r, line 200

[Jean-Romain]

Hi,

Please check and report your:

• R version (mine is 3.2.2)
• data.table package version (mine is 1.9.6)

I found 3 references with this error: here joey711/phyloseq#564 and here http://stackoverflow.com/questions/31142280/unusual-error-in-setkeyv-with-data-table and here http://grokbase.com/t/r/r-help/156m0j2nfm/r-data-table-package-problem. It seems that the problem comes from the R version.

If updating R solves this issue I will change the DESCRIPTION file.

[ptompalski]

R version 3.2.2
data.table 1.9.6

After updating R to 3.2.4 problem does not occur.
This is strange and I am not sure what causes that.

[Jean-Romain]

If the problem comes from R or data.table, it's out of my skills.

Here with the same version than mine (linux) and the same than another user (windows) you have the problem. You updated to a greater version than the one we are using. I can't modify the DESCRIPTION file to impose a greater version than mine because it works on 3.2.2. The problem comes from somewhere deeper than what I can investigate.

[ptompalski]

I will try this on several other machines and report back.

Edit:
works on R 3.2.1, dt 1.9.6
works on R 3.1.2, dt 1.9.6
works on R 3.1.2, dt 1.9.6

all on windows platforms.

Hopefully this was a single issue on my laptop.

[Jean-Romain]

Works also on R 3.2.2 on Windows platform on a friend's computer. We made 7 tests, 6 on Windows, 1 on Linux, on 4 versions of R. I can maybe make more tests on Linux on my other computers but if it's not a single issue it will be very very hard to reproduce it and then, to solve it.

Thank you for your tests. I will consider this bug fixed because we cannot reproduce it. It seems to be specific to something else than pure code.

[ptompalski]

Agree. If it happens again, then we can investigate it more.