search

idinov / informatics

Automatically exported from code.google.com/p/informatics
0 stars 0 forks source link

CNV3 (CNVseq) summary issues #14

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
-CNV3 (CNVseq):

The last discussion we made was:

"(3) CNV3 (CNVseq path): I searched for the Rexe (bc I wanted to avoid bugging 
you anymore but I couldn't find it), so just to keep in mind when you have time.

·        I’m sorry I confused you yesterday! I think Rexe is just a pointer 
to the /uci/fgene location of the R executable. I believe (on fgene1) this 
location may be: /raid/programs/linux/R-2.8.1/bin/R"

I run the attached pipe, having an error in the cnvseq module, no error stream 
but if I run the command line I have an error about the Rexe:

/projects1/test_pipeline/NEW_ORGANIZATION/scripts/cnv-seq.pl --test 
/usr/pl_cache/pipeline/2011August04_16h10m56s535ms/ExtractHits_2.OutputHits-1.hi
ts --ref 
/usr/pl_cache/pipeline/2011August04_16h10m56s535ms/ExtractHits_1.OutputHits-1.hi
ts --genome human --Rexe /applications/R-2.11.1/bin/exec/R 
/usr/pl_cache/pipeline/2011August04_16h10m56s535ms/CNV-Seq_1.OutputCountFile-1.c
ount 
/usr/pl_cache/pipeline/2011August04_16h10m56s535ms/CNV-Seq_1.OutputCNVFile-1.cnv
genome size used for calculation is 3253037807
/usr/pl_cache/pipeline/2011August04_16h10m56s535ms/ExtractHits_2.OutputHits-1.hi
ts: 48477750 reads
/usr/pl_cache/pipeline/2011August04_16h10m56s535ms/ExtractHits_1.OutputHits-1.hi
ts: 34426365 reads
/applications/R-2.11.1/bin/exec/R: error while loading shared libraries: 
libRblas.so: cannot open shared object file: No such file or directory

 Error: can not find program /applications/R-2.11.1/bin/exec/R at /projects1/test_pipeline/NEW_ORGANIZATION/scripts/cnv-seq.pl line 123.

...so the problem is always the path to Rexe, even if I put the path that also 
you suggested and that make more sense also to me.

I don't find any issue about CNV3 on the tracker, so hopefully I anot misisng 
any piece of info.

Original issue reported on code.google.com by federica...@gmail.com on 4 Aug 2011 at 11:30

Attachments:

GoogleCodeExporter commented 9 years ago 
It looks like you are missing the library libRblas.so. I would say try 
installing it and see if the problem persists.

Original comment by alexge...@gmail.com on 5 Aug 2011 at 6:57

GoogleCodeExporter commented 9 years ago 
Yes, I am trying to do it.

Federica

Original comment by federica...@gmail.com on 8 Aug 2011 at 11:48

GoogleCodeExporter commented 9 years ago 
I found the missing library, it's at /applications/R-2.11.1/lib/libRblas.so

I added it to the library path:

export LD_LIBRARY_PATH=/applications/R-2.11.1/lib:$LD_LIBRARY_PATH

and that particular error is gone. We need to do this in all scripts before 
running the R executable.

but now there are other errors, so I'm debugging those as we speak.

Alex

Original comment by alexge...@gmail.com on 12 Aug 2011 at 9:16

GoogleCodeExporter commented 9 years ago 
Hi,

thank you for your help! SO, I added this path to my .bashrc file, my bash 
profile and so on...but still I have a pending question. Probably is a naive 
thing. I am using the local pipeline client right? And I am connected to 
fgene3..but I am connected in bash or tsch? Because I am having the same error:

/applications/R-2.11.1/bin/exec/R: error while loading shared libraries: 
libRblas.so: cannot open shared object file: No such file or directory

 Error:
can not find program /applications/R-2.11.1/bin/exec/R at 
/projects1/test_pipeline/NEW_ORGANIZATION/scripts/cnv-seq.pl line 124.

...so I was asking myself if I have to change something also in my  .cshrc file.

Fed

Original comment by federica...@gmail.com on 15 Aug 2011 at 7:43

GoogleCodeExporter commented 9 years ago 
To be safe, add it to .cshrc and .tcshrc as well:

setenv LD_LIBRARY_PATH /applications/R-2.11.1/lib:${LD_LIBRARY_PATH}

When i login, it reads .cshrc, not .bashrc. Have you tried this? Does it still 
not work?

Original comment by alexge...@gmail.com on 17 Aug 2011 at 9:24

GoogleCodeExporter commented 9 years ago 
ok, I have modified also the .cshrc... actually I don't have a .tcshrc.....BTW 
I have the same error again:

/applications/R-2.11.1/bin/exec/R: error while loading shared libraries: 
libRblas.so: cannot open shared object file: No such file or directory

 Error:
can not find program /applications/R-2.11.1/bin/exec/R at 
/projects1/test_pipeline/NEW_ORGANIZATION/scripts/cnv-seq.pl line 124.

Fe

Original comment by federica...@gmail.com on 18 Aug 2011 at 5:16

GoogleCodeExporter commented 9 years ago 
Oh sorry I misunderstood. If you run this program from a shell session (not the 
pipeline client), the environment variables will be set because .cshrc is read 
when you log into the shell.

From pipeline, we need to set the environment variables through a wrapper 
script. I've done that ( see attached). But again, there are new errors that 
I'm working on right now. I'll update you with those soon.

Alex

Original comment by alexge...@gmail.com on 18 Aug 2011 at 6:04

Attachments:

GoogleCodeExporter commented 9 years ago 
Oki, thank you so much!

fed

Original comment by federica...@gmail.com on 18 Aug 2011 at 6:15

GoogleCodeExporter commented 9 years ago 
Yep, I have this one (donno if it can help):

Illegal division by zero at 
/projects1/test_pipeline/NEW_ORGANIZATION/scripts/cnv-seq.pl line 168, <TEST> 
line 1207.

Fed

Original comment by federica...@gmail.com on 18 Aug 2011 at 6:34

Attachments:

GoogleCodeExporter commented 9 years ago 
I tried again after the update of teh pipeline, but still I have the same 
error...

federica

Original comment by federica...@gmail.com on 6 Sep 2011 at 8:01

GoogleCodeExporter commented 9 years ago 
I'm still working on debugging this, it's slow going though because my 
connection to fgene1 gets dropped frequently.

Original comment by alexge...@gmail.com on 6 Sep 2011 at 8:37

GoogleCodeExporter commented 9 years ago 
yep, I had the same problem from off campus when I was in Italy...do you want 
me to signal this problem to our system admin? Can I do something to help you 
out?

Fed

Original comment by federica...@gmail.com on 6 Sep 2011 at 9:36

GoogleCodeExporter commented 9 years ago 
We will contact him.

Original comment by alexge...@gmail.com on 7 Sep 2011 at 6:33

GoogleCodeExporter commented 9 years ago 
I am writing some technical emails to Bernard, so do you want me to write him 
also about this one?

fed

Original comment by federica...@gmail.com on 9 Sep 2011 at 5:34

GoogleCodeExporter commented 9 years ago 
sure, you can mention the occasional issues connecting to the server remotely 
if you want. Something tells me he already knows though :)

Original comment by alexge...@gmail.com on 9 Sep 2011 at 5:53

GoogleCodeExporter commented 9 years ago 
Now there are not errors...but it's running since 5 days, just to keep track of 
it.

fed

Original comment by federica...@gmail.com on 15 Sep 2011 at 5:37

Attachments:

GoogleCodeExporter commented 9 years ago 
same here, been going for almost two days now. Does this seem reasonable to 
you?

Original comment by alexge...@gmail.com on 15 Sep 2011 at 6:01

GoogleCodeExporter commented 9 years ago 
Actually no, it runs super fast usually...

Original comment by federica...@gmail.com on 15 Sep 2011 at 6:04

GoogleCodeExporter commented 9 years ago 
Ok, I assume this is taking too long. I've been messing with the source for a 
while to debug, and although I don't think I did anything intrusive, just in 
case, I did a clean installation of the cnv-seq package to reverse all my 
changes and see if we still get an infinite loop. If the failures from before 
were a result of the strange server issues we were having, this might fix a lot.

I'm running it now, I'll let you know.
Alex

Original comment by alexge...@gmail.com on 19 Sep 2011 at 6:27

Attachments:

GoogleCodeExporter commented 9 years ago 
Thanks, I am doing the same right away.

fed

Original comment by federica...@gmail.com on 19 Sep 2011 at 6:46

GoogleCodeExporter commented 9 years ago 
Hi,

I ran the module, and this time the CNVseq run was ok. BTW, there is a problem 
in the CNV-R connection on the system...as the cnv library is not present in 
the server (see pasted error and pics:

  File "<string>", line 1
    print len(''chromosome=2,from=140036061,to=144238634''.split())
                         ^
SyntaxError: invalid syntax
/projects1/test_pipeline/NEW_ORGANIZATION/scripts/cnv-seq-R.sh: line 13: [: 
-gt: unary operator expected
Error in library(cnv) : there is no package called 'cnv'
Execution halted
Error in library(cnv) : there is no package called 'cnv'
Execution halted
cp: cannot stat `*.pdf': No such file or directory

Actually I have installed the package on my local mac with this procedure:

>  install.packages("cnv")
> ./R CMD INSTALL /applications/CNVseq/cnv-seq/cnv/

Have you installed this package on your side? I remember that we had some 
issues with this package on fgene...and here I think is the reason of the 
failure..

Fed

Original comment by federica...@gmail.com on 20 Sep 2011 at 12:09

Attachments:

GoogleCodeExporter commented 9 years ago 
More in detail here is what I did on my Unix local machine:

CNVseq:  LOCAL MAC

-go to the folder with the cnv/folder and:

Federica-Torris-iMac:cnv_seq federicatorri$ R CMD INSTALL cnv/

-from GUI install the necessary library

-then run

./cnv-seq.pl --test JLK-227.hits --ref JAM-230.hits --genome human

I remember that on our server I was facing a lot of problems and that also with 
Bernard finally the issue was not fixed..and then I ha dto run everything on my 
local machine.

Sorry to bug you and hope it may help to see what happens on your Linux server 
with the same issue.

Fed

Original comment by federica...@gmail.com on 20 Sep 2011 at 12:18

GoogleCodeExporter commented 9 years ago 
Oh, excellent. I'm past the CNV-seq step as well. Fixing CNV-R shouldn't be too 
complicated, looks like a simple python syntax error, probably my bad. I'll get 
back to you soon.

Alex

Original comment by alexge...@gmail.com on 20 Sep 2011 at 2:53

GoogleCodeExporter commented 9 years ago 
Hi Alex,

I want only to add something as I think I haven't explained everything properly 
(sorry to re-bug you if I was clear instead :) ). 

I fixed the problem going "around it" and using my local mac but CNVseq must 
run on a linux server given the great amount of data. SO I explained to you how 
I fixed it "locally" to see if you also can help me out (looking at how you 
deal with it on your server) in understanding how the problem can be fixed also 
on fgene, bc I have not been able to fix it to date by myself.

Thanks,

Fede

Original comment by federica...@gmail.com on 20 Sep 2011 at 5:18

GoogleCodeExporter commented 9 years ago 
Yeah, so the CNV R library is missing on fgene1. We went through this on our 
servers as well and I think installing CNV fixed the problem (looks like it 
fixed the problem on your local machine). Should I email Bernard about 
installing this package on fgene?

Is it just a simple `R CMD INSTALL /applications/CNVseq/cnv-seq/cnv/` ?

Original comment by alexge...@gmail.com on 20 Sep 2011 at 6:35

GoogleCodeExporter commented 9 years ago 
Yep, I will email him about it cc-ing you. Please let me know also if you hear 
from him abotu abyss or if I have to ask him again something.

fed

Original comment by federica...@gmail.com on 20 Sep 2011 at 6:39

GoogleCodeExporter commented 9 years ago 
Oh no sorry, ABYSS is with Sam...:)

Original comment by federica...@gmail.com on 20 Sep 2011 at 6:48

GoogleCodeExporter commented 9 years ago 
Hi Alex,

bernard pasted my this:

[root@fgene2 bin]# ./R CMD INSTALL /applications/CNVseq/cnv-seq/cnv/
* installing to library â/applications/R-2.11.1/libraryâ
* installing *source* package âcnvâ ...
** R
** preparing package for lazy loading
Loading required package: reshape
Loading required package: plyr

Attaching package: 'reshape'

The following object(s) are masked from 'package:plyr':

   round_any

Loading required package: grid
Loading required package: proto
** help
No man pages found in package  âcnvâ
*** installing help indices
** building package indices ...
** testing if installed package can be loaded

* DONE (cnv)
[root@fgene2 bin]# 

I think now is ok also on our system..I will give a try.

I'll let you know.

Tks

Fed

Original comment by federica...@gmail.com on 20 Sep 2011 at 9:26

GoogleCodeExporter commented 9 years ago 
The problem now is that CNV-seq was generating an unpredictable output .cnv 
file, so the pipeline module couldn't pass it to CNVR and CNVR was failing 
because of the missing file. I fixed CNV-seq to properly extract the output 
filename. If you have an active session of this workflow right now, just 
restart the CNV-seq module. I'm running it now.

btw the last time i ran cnv-seq it ran for several hours, but it completed. the 
current run is taking over an hour as well. should we be concerned about this? 

alex

Original comment by alexge...@gmail.com on 21 Sep 2011 at 7:54

GoogleCodeExporter commented 9 years ago 
I have restarted it right now. About the timing I remember that on my local 
machine was taking alot as well, even if less than days like that weird past 
run..so I think that everything is ok..I am eager to see the output files, than 
I;ll be able to tell you that is working ok.

Tks!

Fed

Original comment by federica...@gmail.com on 21 Sep 2011 at 8:46

GoogleCodeExporter commented 9 years ago 
Hi, the run was successful this time...everything's green! It's great! 

There is only something I need to change and discuss with you.

(1) First of all there is a duplication of the outputs:

[ftorri@fgene2 2011September21_13h36m46s971ms]$ ll -h
total 3.4G
-rw-rw-rw- 1 pipeline pipeline  155 Sep 21 19:37 CNV-R_1.OutputDataDescription-1
-rw-rw-rw- 1 pipeline pipeline 1.1K Sep 21 19:37 CNV-R_1.OutputPlot-1.pdf
-rw-rw-rw- 1 pipeline pipeline 111M Sep 21 19:33 CNV-Seq_1.OutputCNVFile-1.cnv
-rw-rw-rw- 1 pipeline pipeline 1.2G Sep 21 13:56 ExtractHits_1.OutputHits-1.hits
-rw-rw-rw- 1 pipeline pipeline 2.0G Sep 21 14:10 ExtractHits_2.OutputHits-1.hits
-rw-rw-rw- 1 pipeline pipeline 111M Sep 21 19:33 
ExtractHits_2.OutputHits-1.hits-vs-ExtractHits_1.OutputHits-1.hits.log2-0.6.pval
ue-0.001.minw-4.cnv
-rw-rw-rw- 1 pipeline pipeline  29M Sep 21 14:26 
ExtractHits_2.OutputHits-1.hits-vs-ExtractHits_1.OutputHits-1.hits.log2-0.6.pval
ue-0.001.minw-4.count
-rw-rw-rw- 1 pipeline pipeline 1.1K Sep 21 19:37 Rplots.pdf
drwxrwxrwx 2 pipeline pipeline 4.0K Sep 21 19:33 streams

The CNV-Seq_1.OutputCNVFile-1.cnv and the 
ExtractHits_2.OutputHits-1.hits-vs-ExtractHits_1.OutputHits-1.hits.log2-0.6.pval
ue-0.001.minw-4.cnv are the SAME thing. So it's better that the .cnv file has 
the name 
:ExtractHits_2.OutputHits-1.hits-vs-ExtractHits_1.OutputHits-1.hits.log2-0.6.pva
lue-0.001.minw-4.cnv (bc incorporates the information about the comparison 
between cases and controls). I think the issue is that with the pipeline we 
have specified the output name as CNV-Seq_1.OutputCNVFile-1.cnv, while the 
software per se is producing by default an output with the other name also.

(2) There is also something related to the logic of the last step (CNV-R). The 
pdf plot is empty and  the reason is that for the testing you use the interval 
chromosome=2,from=140036061,to=144238634. We are demonstrating that the command 
line has worked, and this is awesome!  BTW the plot is empty bc probably in 
this patient we don't have a deletion in this genomic interval.  I describe you 
the logical flow of what the user might want to do.

-automatically EXPORT ALL the picture of the CNVs in all the genome

-automatically EXPORT ALL the cnv in a specific chromosome

-EXPORT ONLY some CNVs he found interesting after having examinated the .cnv 
file. So here the user might need to stop the pipeline run at the CNVseq step, 
download the .cnv file, give a look at that and once he has a list of the 
important events he would need to go back to the same workflow and run the 
CNV-R step with the coordinates or the CNVid he wants to plot (in the .cnv file 
the user can in fact find either the coordinates of the CNV or a specific ID 
that is a number).

I am discovering in these days these new options of outputting. So in summary, 
this  a great success because this module was really hard to make it working, 
tks for your help. I am attaching the updated version of the documentation 
where I tried to explain everything. I am really interested in knowing what do 
you think about those changes: do you think are too much complex and would 
require too much work? 

The other output of CNV-R module (production of the data description file) is 
ok and perfect like that.

Thanks,

Federica

Original comment by federica...@gmail.com on 22 Sep 2011 at 6:03

Attachments:

GoogleCodeExporter commented 9 years ago 
Yes, this completed for me too.

(1) This was a somewhat quick fix on my part. The reason is that the long 
filename 
(ExtractHits_2.OutputHits-1.hits-vs-ExtractHits_1.OutputHits-1.hits.log2-0.6.pva
lue-0.001.minw-4.cnv) is generated by CNV-seq at runtime, and not specified on 
the command line beforehand. So Pipeline doesn't know what this filename will 
be, and therefore can't access it. So as a workaround, after cnv-seq completes, 
there is a cp step that copies any *.cnv file in the working directory to a 
pipeline-generated filename (CNV-Seq_1.OutputCNVFile-1.cnv). Otherwise, the 
next module won't be able to find the .cnv file.

If this is going to be an issue, I'm sure I could figure out a way for pipeline 
to assemble the filename at runtime based on parameters. Let me know if you 
think this is worthwhile.

(2) I just copied those arguments exactly from whatever documentation I was 
reading at the time, so I'm not surprised it's irrelevant. 

For the first case (full genome): is this just a matter of leaving out the 
chromosome= and from=,to= flags? Is this the default behavior? Otherwise how do 
you specify graphing the entire genome?

Second and third case: These will need to be run in separate pipeline 
workflows. The process would involve completing the workflow, examining the 
output, then doing a "copy with input" on the CNV-R module, pasting it into a 
blank workflow, and running it with the desired new parameters. 

The other options is to send the output of CNV-seq to several parallel 
instances of CNV-R, each running a different set of parameters. You would have 
to have a general idea of what areas you are interested in though. I can't 
think of a more elegant solution at the moment.

I'm going to test the CNV-R module further to make sure it can generate a 
proper pdf. Let me know if anything is unclear, or anything else I can put 
together.

Alex

Original comment by alexge...@gmail.com on 22 Sep 2011 at 7:01

GoogleCodeExporter commented 9 years ago 
Thank you for your quick reply! I really appreciate your suggestions.

(1) No I mean for the doubled name we can by now leave it like that (describing 
btw what you told me, so the user won't be surprised).

(2) Actually I was testing the command lines but they are not working well ( I 
mean the whole chr and whole genome)..so at this point maybe is better to make 
things simple and to allow the user to export the regions he wants..that brings 
us to the last point.

The idea to split the two workflows is the best one in my opinion ...maybe I 
would keep them in the same canva like in the pic I attach here. The only thing 
I would like to ask you is where the user can specify the genomic coordinates 
or the CNVid he wants to visualize. I saw that in the parameter is explained an 
example of the posisble arguments...but where is phisically the command line? 
In any script? The thing that may be useful is a place where the user can enter 
the parameters choosing the genomic location of the event he wants to visualize.

The command line may ask for regions: plot.cnv(data, chromosome=2, 
from=140036061, to=144238634)

or CNvid: plot.cnv(data, CNV=4, upstream=4e+6, downstream=4e+6) 

For the testing you are doing I suggest you to use the region chromosome 3 from 
    78154481    to  78160016 where actually there is a CNV in the tested case.

Fede

Original comment by federica...@gmail.com on 22 Sep 2011 at 7:57

Attachments:

GoogleCodeExporter commented 9 years ago 
So all the arguments to plot.cnv (besides "data") are specified in the "Plot 
Arguments" input parameter of the CNV-R module. You simply list the arguments 
as they should be called in R. For example, if you want to call plot.cnv(data, 
chromosome=2, from=140036061, to=144238634), you simply enter the string:

chromosome=2, from=140036061, to=144238634

The pipeline will automatically escape the '=' and whitespace, so you don't 
need to worry about formatting, as long as they are specified with the proper R 
syntax. (the description of this parameter says no spaces, but go ahead and use 
spaces if you want, I'll remove this warning).

That being said, I just ran this with these parameters that you suggested and 
got a failure:

Error in eval(expr, envir, enclos) : object 'position' not found
Calls: print ... lapply -> is.vector -> lapply -> FUN -> eval -> eval
In addition: Warning messages:
1: In max(na.omit(data$p.value)) :
  no non-missing arguments to max; returning -Inf
2: In match.fun(get(".transform", .))(values) : NaNs produced
Execution halted

I'll investigate further, but does this tell you anything off the bat?

Alex

Original comment by alexge...@gmail.com on 22 Sep 2011 at 10:12

GoogleCodeExporter commented 9 years ago 
Hi,

thanks. I am running it on another region I foudn to be affected by a deletion 
in the sample. ("22"    49577649    49583184) I don't have idea about the reason of 
this error you are getting, I have never had it..lets see what comes out with 
my run.

fed

Original comment by federica...@gmail.com on 22 Sep 2011 at 10:51

GoogleCodeExporter commented 9 years ago 
So this works for me now. Attached the output of my previous run with 
chromosome=2,from=140036061,to=144238634 (same as before)

Original comment by alexge...@gmail.com on 26 Sep 2011 at 6:32

Attachments:

GoogleCodeExporter commented 9 years ago 
Hi,

I am running CNVseq on a region with CNV..it is taking a lot (3 days ) and I 
think it is stalling.I am n ot having nay error though. The module si exaclty 
the same, I have only changed the interval as you can see from the shot. Have 
you any news about your run?

Fed

Original comment by federica...@gmail.com on 26 Sep 2011 at 6:34

Attachments:

GoogleCodeExporter commented 9 years ago 
just posted about this.  CNV-seq tool 6 hours for me. Is the output stream 
still reporting? If it's still doing work, it will print something like [1] 
"cnv_id:  33634  of  33641" with the numbers incrementing. Otherwise, i would 
restart it.

And if you didn't notice my post, check it out.

Original comment by alexge...@gmail.com on 26 Sep 2011 at 6:41

GoogleCodeExporter commented 9 years ago 
Hi,

I have already restarted the module this morning as the output stream was not 
reporting anymore. I hope it will end with a not daily timing...your plot is 
perfect. Can you copy the working pipe in the library? SO I'm sure to have your 
working copy.

tks

Federica

Original comment by federica...@gmail.com on 26 Sep 2011 at 9:32

GoogleCodeExporter commented 9 years ago 
here it is. I don't have permissions to add to the pipeline library. i can put 
it somewhere on fgene if you want though.

Original comment by alexge...@gmail.com on 26 Sep 2011 at 9:41

Attachments:

GoogleCodeExporter commented 9 years ago 
I'll do that, no problem ...many thanks!

fed

Original comment by federica...@gmail.com on 26 Sep 2011 at 9:49

GoogleCodeExporter commented 9 years ago 
Perfect, it ran for me as well with the plot.

Fed

Original comment by federica...@gmail.com on 28 Sep 2011 at 12:16

GoogleCodeExporter commented 9 years ago 
Hi, when using the .cnv output from cnv-seq.pl and trying to plot with 
library(cnv) from cnv-seq in R there appears an error like this. I am upset 
because I have generated 7 files and some of them are working fine and some not 
and they are generated in the same way. Here I attach the error in R cran: 

> library(cnv)
> data <- read.delim("my_data.cnv")
> cnv.print(data)
cnv     chromosome      start   end     size    log2    p.value
CNVR_1  chr     Inf     -Inf    -Inf
CNVR_0  chrY    chrX    chrMT   chr1    chr2    chr3    chr4    chr5    chr6    
chr7    chr8
chr9    chr10   chr11   chr12   chr13   chr14   chr15   chr16   chr17   chr18   
398
315321616       315321219       NA      NA
Warning messages:
1: In min(sub$start) : no non-missing arguments to min; returning Inf
2: In min(sub$position) : no non-missing arguments to min; returning Inf
3: In max(sub$end) : no non-missing arguments to max; returning -Inf
4: In max(sub$position) : no non-missing arguments to max; returning -Inf

Thank you very much in advance!

Original comment by anna.pui...@gmail.com on 25 Mar 2014 at 11:34

GoogleCodeExporter commented 9 years ago 
Hello, Anna

I had an exactly same problem with you.. 
Did you solve this problem?? I also could not use cnv.print.. (same error with 
you) but, some of dataset works fine.. ;( too weird.. 

Somebody any helps??

Original comment by younhee...@gmail.com on 3 Jun 2014 at 1:03