azofeifa / FStitch

An algorithm to detect nascent RNA transcription in GRO-seq and ChIP-seq
7 stars 6 forks source link

Core dump #1

Closed gatozee closed 8 years ago

gatozee commented 8 years ago

The program core dumps when I run it on a training set. If you need, I can provide the input files.

azofeifa commented 8 years ago

Yes that would be great! Thank you very much! cheers, Joey

On Tue, Nov 24, 2015 at 11:32 AM, gatozee notifications@github.com wrote:

The program core dumps when I run it on a training set. If you need, I can provide the input files.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1.

gatozee commented 8 years ago

The command I used was: FStitch -i example.bedgraph -j trainingSet.txt -o example.param

trainingSet.txt and example.bedgraph can be downloaded from https://www.dropbox.com/sh/cdvq7boepeixvn3/AADniHGetsq9Dm6IarOo7MOda?dl=0

Thank you!

On Tue, Nov 24, 2015 at 1:46 PM, Joey Azofeifa notifications@github.com wrote:

Yes that would be great! Thank you very much! cheers, Joey

On Tue, Nov 24, 2015 at 11:32 AM, gatozee notifications@github.com wrote:

The program core dumps when I run it on a training set. If you need, I can provide the input files.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159369162.

azofeifa commented 8 years ago

Hi so I think I found the source of one possible error, although I was unable to recapitulate the core-dump I did see that I was not checking for carriage returns (are you working on a windows machine?), so I fixed that. At any rate, please re-clone from my github site and run with the -v option (verbose output) and if the error persists on these input files send me the standard output. Be sure to recompile.

Thanks very much and I am sorry for this inconvenience.

cheers, Joey

On Tue, Nov 24, 2015 at 11:54 AM, gatozee notifications@github.com wrote:

The command I used was: FStitch -i example.bedgraph -j trainingSet.txt -o example.param

trainingSet.txt and example.bedgraph can be downloaded from https://www.dropbox.com/sh/cdvq7boepeixvn3/AADniHGetsq9Dm6IarOo7MOda?dl=0

Thank you!

On Tue, Nov 24, 2015 at 1:46 PM, Joey Azofeifa notifications@github.com wrote:

Yes that would be great! Thank you very much! cheers, Joey

On Tue, Nov 24, 2015 at 11:32 AM, gatozee notifications@github.com wrote:

The program core dumps when I run it on a training set. If you need, I can provide the input files.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159369162.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159371323.

gatozee commented 8 years ago

Thank you! I recompiled the new version. But it still core dumped before -v showed anything. I'm on Linux--just the traningset file is from someone using Windows.

It compiled successfully:

Compiled main Compiled BaumWelch Compiled NewtonsMethod Compiled read_in_parameters Compiled read Compiled main_train Compiled grabTrainingExamples Compiled interval_tree Compiled split Compiled write Compiled main_segment Compiled viterbi

Compiled validate

Sucessfully Compiled

azofeifa commented 8 years ago

interesting! Well, let me look at it again right now and will get back to you, sorry again for the inconvenience. cheers, Joey

On Wed, Nov 25, 2015 at 9:50 AM, gatozee notifications@github.com wrote:

Thank you! I recompiled the new version. But it still core dumped before -v showed anything. I'm on Linux--just the traningset file is from someone using Windows. It compiled successfully:

Compiled main Compiled BaumWelch Compiled NewtonsMethod Compiled read_in_parameters Compiled read Compiled main_train Compiled grabTrainingExamples Compiled interval_tree Compiled split Compiled write Compiled main_segment Compiled viterbi Compiled validate

Sucessfully Compiled

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159669184.

azofeifa commented 8 years ago

So I see in your command

FStitch -i example.bedgraph -j trainingSet.txt -o example.param

that you are missing

"train"

try:

FStitch train -i example.bedgraph -j trainingSet.txt -o example.param

and let me know if you get anymore errors. It still shouldn't be core dumping sense I check for that but lets see what happens anyway.

cheers, Joey

On Wed, Nov 25, 2015 at 9:56 AM, Joey Azofeifa <joseph.azofeifa@colorado.edu

wrote:

interesting! Well, let me look at it again right now and will get back to you, sorry again for the inconvenience. cheers, Joey

On Wed, Nov 25, 2015 at 9:50 AM, gatozee notifications@github.com wrote:

Thank you! I recompiled the new version. But it still core dumped before -v showed anything. I'm on Linux--just the traningset file is from someone using Windows. It compiled successfully:

Compiled main Compiled BaumWelch Compiled NewtonsMethod Compiled read_in_parameters Compiled read Compiled main_train Compiled grabTrainingExamples Compiled interval_tree Compiled split Compiled write Compiled main_segment Compiled viterbi Compiled validate

Sucessfully Compiled

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159669184.

gatozee commented 8 years ago

Sorry, I did use "train":

FStitch train -v -i example.bedgraph -j trainingSet.txt -o example.param Segmentation fault (core dumped)

uname -a Linux k2 4.2.0-16-generic #19-Ubuntu SMP Thu Oct 8 15:35:06 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux

On Wed, Nov 25, 2015 at 12:05 PM, Joey Azofeifa notifications@github.com wrote:

So I see in your command

FStitch -i example.bedgraph -j trainingSet.txt -o example.param

that you are missing

"train"

try:

FStitch train -i example.bedgraph -j trainingSet.txt -o example.param

and let me know if you get anymore errors. It still shouldn't be core dumping sense I check for that but lets see what happens anyway.

cheers, Joey

On Wed, Nov 25, 2015 at 9:56 AM, Joey Azofeifa < joseph.azofeifa@colorado.edu

wrote:

interesting! Well, let me look at it again right now and will get back to you, sorry again for the inconvenience. cheers, Joey

On Wed, Nov 25, 2015 at 9:50 AM, gatozee notifications@github.com wrote:

Thank you! I recompiled the new version. But it still core dumped before -v showed anything. I'm on Linux--just the traningset file is from someone using Windows. It compiled successfully:

Compiled main Compiled BaumWelch Compiled NewtonsMethod Compiled read_in_parameters Compiled read Compiled main_train Compiled grabTrainingExamples Compiled interval_tree Compiled split Compiled write Compiled main_segment Compiled viterbi Compiled validate

Sucessfully Compiled

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159669184.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159674152.

gatozee commented 8 years ago

BTW, even running the program itself core dumps:

FStitch please specify either: train or segment exiting... Segmentation fault (core dumped)

azofeifa commented 8 years ago

okay, I know this is a serious pain, I can't seem to get the error on my end (tried it on a linux and mac) but I did make some updates that I think might be the source of the problem for getting a seg fault from just running the program without anything. Again please re-clone and recompile.

Real quick: what GNU compiler version are you using?

cheers, Joey

On Wed, Nov 25, 2015 at 11:01 AM, gatozee notifications@github.com wrote:

BTW, even running the program itself core dumps:

FStitch please specify either: train or segment exiting... Segmentation fault (core dumped)

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159688758.

gatozee commented 8 years ago

gcc 5.2.1

On Wed, Nov 25, 2015 at 1:10 PM, Joey Azofeifa notifications@github.com wrote:

okay, I know this is a serious pain, I can't seem to get the error on my end (tried it on a linux and mac) but I did make some updates that I think might be the source of the problem for getting a seg fault from just running the program without anything. Again please re-clone and recompile.

Real quick: what GNU compiler version are you using?

cheers, Joey

On Wed, Nov 25, 2015 at 11:01 AM, gatozee notifications@github.com wrote:

BTW, even running the program itself core dumps:

FStitch please specify either: train or segment exiting... Segmentation fault (core dumped)

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159688758.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159690933.

azofeifa commented 8 years ago

okay that should be good, this is a fairly annoying process sense I can't get the error over here, but I appreciate your patience this only helps make the program more stable, so thanks!

On Wed, Nov 25, 2015 at 11:12 AM, gatozee notifications@github.com wrote:

gcc 5.2.1

On Wed, Nov 25, 2015 at 1:10 PM, Joey Azofeifa notifications@github.com wrote:

okay, I know this is a serious pain, I can't seem to get the error on my end (tried it on a linux and mac) but I did make some updates that I think might be the source of the problem for getting a seg fault from just running the program without anything. Again please re-clone and recompile.

Real quick: what GNU compiler version are you using?

cheers, Joey

On Wed, Nov 25, 2015 at 11:01 AM, gatozee notifications@github.com wrote:

BTW, even running the program itself core dumps:

FStitch please specify either: train or segment exiting... Segmentation fault (core dumped)

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159688758.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159690933.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-159691303.

gatozee commented 8 years ago

Now it runs without any segfault. Thanks a lot!

azofeifa commented 8 years ago

glad it worked out! let me know if anything else comes up!

gatozee commented 8 years ago

Can I ask a general question about FStiitch here? Do you recommend creating a specific training set for each sample or using the training set from one sample for a batch of samples? Thanks!

On Wed, Nov 25, 2015 at 5:17 PM, Joey Azofeifa notifications@github.com wrote:

Closed #1 https://github.com/azofeifa/FStitch/issues/1.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#event-475023653.

azofeifa commented 8 years ago

Thats a great question! So in general I would say making a batch training set is completely reasonable as long as those training examples are consistent across samples. You wouldn't want a situation where after some experiment an inactive region becomes an actively transcribed region. So as long as your training examples are behaving consistently across sample types than I think it will work fine.

On Fri, Dec 4, 2015 at 12:11 PM, gatozee notifications@github.com wrote:

Can I ask a general question about FStiitch here? Do you recommend creating a specific training set for each sample or using the training set from one sample for a batch of samples? Thanks!

On Wed, Nov 25, 2015 at 5:17 PM, Joey Azofeifa notifications@github.com wrote:

Closed #1 https://github.com/azofeifa/FStitch/issues/1.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#event-475023653.

— Reply to this email directly or view it on GitHub https://github.com/azofeifa/FStitch/issues/1#issuecomment-162055894.