Question

Problem in extracting a row from matrix in rhdf5

0

Entering edit mode

filipe • 0

@filipe-13062

Last seen 7.6 years ago

Hello,

I'm having trouble extracting a specific row from a big matrix using the rhdf5 package. It works if I give it an integer as the row index, but not if I assign that integer to a variable and then use that variable when subsetting. Please see the minimal example below for the error and session info.

If anyone has encountered this and/or can spot an error in my code that would be great! Thanks in advance!

Cheers,

Filipe

>
> library(rhdf5)
>
> h5f <- H5Fopen("/path/to/file.h5")
>
> big_matrix <- h5f&"big_matrix"
>
> # This works
> extracted_row <- big_matrix[293857, ]

>
> # This doesn't
> row_index <- 293857
> big_matrix[row_index, ]
Error in as.integer(index[[i]]) :
cannot coerce type 'symbol' to vector of type 'integer'
Error in H5Dread(h5dataset = h5dataset, h5spaceFile = h5spaceFile, h5spaceMem = h5spaceMem, :
HDF5. Dataset. Read failed.
Error in sort.int(x, na.last = na.last, decreasing = decreasing, ...) :
'x' must be atomic
>
> sessionInfo()
R version 3.3.1 (2016-06-21)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Red Hat Enterprise Linux Server release 6.8 (Santiago)

locale:
[1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
[3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
[5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
[7] LC_PAPER=en_US.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] stats graphics grDevices utils datasets methods base

other attached packages:
[1] rhdf5_2.18.0

loaded via a namespace (and not attached):
[1] zlibbioc_1.20.0 tools_3.3.1 packrat_0.4.8-1
>

rhdf5 • 2.5k views

ADD COMMENT • link updated 7.6 years ago by Mike Smith ★ 6.6k • written 7.7 years ago by filipe • 0

0

Entering edit mode

Update to my initial post: this code still doesn't work using R-3.4.0 and rhdf5_2.20.0. However, it does give a different error message:

> row_index <- 293857
> big_matrix[row_index,]
Error in H5Sselect_index(h5spaceFile, index) :
  index exceeds HDF5-array dimension.
Error in H5Dread(h5dataset = h5dataset, h5spaceFile = h5spaceFile, h5spaceMem = h5spaceMem,  :
  HDF5. Dataset. Read failed.
Error in sort.int(x, na.last = na.last, decreasing = decreasing, ...) :
  'x' must be atomic
>
> sessionInfo()
R version 3.4.0 (2017-04-21)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Red Hat Enterprise Linux Server release 6.4 (Santiago)

ADD REPLY • link 7.7 years ago filipe • 0

0

Entering edit mode

Hi Filipe,

Thanks for reporting this. I've managed to reproduce this with a simple example. I'll look into why it's happening and report back once I've got a solution.

ADD REPLY • link 7.7 years ago Mike Smith ★ 6.6k

score 0 · Answer 1 · 2017-06-13

0

Entering edit mode

Mike Smith ★ 6.6k

@mike-smith

Last seen 9 hours ago

EMBL Heidelberg

This should be fixed in rhdf5 version 2.21.2. I use H5DOpen() to access the dataset, but hopefully this example is demonstrates the fix.

library(rhdf5)

tf <- tempfile(fileext = ".h5")
download.file(url = "http://msmith.de/data/h5tutr_dset.h5",
              destfile = tf,
              quiet = TRUE)
fid <- H5Fopen(tf)
did <- H5Dopen(fid, "dset")

> did[2,]
[1] 0 0 0 0
> 
> row_index <- 2
> did[row_index, ]
[1] 0 0 0 0