[BioC] [Bioc] physical memory dump when using cellHTS 1.4

[Wolfgang Huber]

Dear Russell,

the package cellHTS only uses R code, so we need to figure out which
package that R depends on (or what in R itself) causes the blue screen
of death, and to fix or at least avoid it.

To narrow down the problem:

1. Can you try switching off the image map generation by supplying
"map=FALSE" to the "imageScreen" function. This function is called from
"writeReport" and arguments for it can be supplied via the latter's
"imageScreenArgs" argument.

2. Similarly, try to switch off the tcl/tk progress bar by setting
"progressReport=FALSE".

3. Also try "plotPlateArgs=FALSE"

4. Upgrade to R 2.4.1.
http://www.stats.bris.ac.uk/R

5. If it still fails, can you send us the produced output (as far as the
function got until it crashed), this might help narrow down where/when
it happens.

6. ultimately, I could have a look at your data in private and see
whether I can reproduce and locate the problem.

 Best wishes
  Wolfgang


PS I moved this to the bioconductor list since it is probably more
appropriate.

------------------------------------------------------------------
Wolfgang Huber  EBI/EMBL  Cambridge UK  http://www.ebi.ac.uk/huber


> Hi all,
> I've been using cellHTS on a resonably large dataset.  Everything seems to
> work fine until I try and write the report.
> My laptop seems to be having trouble completing the html output and causes a
> physical memory dump (aka Blue screen of death).
> It seems to be hanging when writing the html for all the mouse
> overs on the Screen-wide heatmap image.
> 
> I'm using R 2.4, and cellHTS 1.4
> 
> any ideas on what's causing this and how I might go about solving the
> problem ?
> 
> cheers
> Russell
>