Page 1 of 6

Datasets

PostPosted: Fri Jan 23, 2015 9:55 pm
by meninonas
Dr. Yoo,

Please find attached the CSV file for the Cell Cycle. I tried running the code by removing the commented line; nevertheless, it still gave the following error:

Code: Select all
-----------------------------------------------------------------------------
                                 ERROR DETAILS
-----------------------------------------------------------------------------
- Banjo                        Bayesian Network Inference with Java Objects -
- Release 2.2.0                                                 15 Apr 2008 -
- Licensed from Duke University                                             -
- Copyright (c) 2005-08 by Alexander J. Hartemink                           -
- All rights reserved                                                       -
-----------------------------------------------------------------------------
- Project:                                               banjo static example
- User:                                                                  demo
- Dataset:                                           33-vars-320-observations
- Notes:                                    static bayesian network inference
-----------------------------------------------------------------------------

[ERROR: Banjo 2.2.0, 2/5/15 10:52:51 PM]
The following info is provided to assist in trouble-shooting:
The text line ' 0' in 'static.mandatory.str' does not contain enough information to specify a variable and its parent info.

(Warning (dev)) The number of supplied variable names (132) needs to match the number of variables (123); Banjo defaults to the variable index numbers instead.

Stack trace info:

edu.duke.cs.banjo.bayesnet.EdgesAsArrayWithCachedStatistics.loadStructureFile(EdgesAsArrayWithCachedStatistics.java:2862)
edu.duke.cs.banjo.bayesnet.EdgesAsArrayWithCachedStatistics.<init>(EdgesAsArrayWithCachedStatistics.java:633)
edu.duke.cs.banjo.bayesnet.BayesNetManager.<init>(BayesNetManager.java:273)
edu.duke.cs.banjo.learner.SearcherSimAnneal.setupSearch(SearcherSimAnneal.java:1092)
edu.duke.cs.banjo.learner.SearcherSimAnneal.<init>(SearcherSimAnneal.java:723)
edu.duke.cs.banjo.application.BanjoThreadHandler.<init>(BanjoThreadHandler.java:115)
edu.duke.cs.banjo.application.Banjo.runMultipleInstances(Banjo.java:200)
edu.duke.cs.banjo.application.Banjo.execute(Banjo.java:150)
edu.duke.cs.banjo.application.Banjo.main(Banjo.java:447)

-----------------------------------------------------------------------------
                           End of error notification
-----------------------------------------------------------------------------

-----------------------------------------------------------------------------
(Final Checkpoint, after post-processing) A final check revealed the following issues that were encountered during Banjo's execution:
-----------------------------------------------------------------------------
(Post-processing) Postprocessing cannot proceed because we can't process the high score network(s).


I'll go over the settings with you tomorrow.

Re: Datasets

PostPosted: Mon Jan 26, 2015 7:30 pm
by meninonas
Dr. Yoo,

I installed Banjo on the Virtual Machine and on the computer and the following error was given to me:

Code: Select all
(Post-processing) Postprocessing cannot proceed because we can't process the high score network(s).

Re: Datasets

PostPosted: Tue Jan 27, 2015 11:45 am
by cwyoo
meninonas wrote:Dr. Yoo,

I installed Banjo on the Virtual Machine and on the computer and the following error was given to me:

Code: Select all
(Post-processing) Postprocessing cannot proceed because we can't process the high score network(s).


Please install banjo in path4 machine.

Re: Datasets

PostPosted: Tue Jan 27, 2015 12:35 pm
by meninonas
Dr. Yoo,

I installed Banjo on the Virtual Machine and on the computer and the following error was given to me:

Code: Select all
-----------------------------------------------------------------------------
(Final Checkpoint, after post-processing) A final check revealed the following issues that were encountered during Banjo's execution:
-----------------------------------------------------------------------------
(Post-processing) Postprocessing cannot proceed because we can't process the high score network(s).

Re: Datasets

PostPosted: Tue Jan 27, 2015 1:58 pm
by meninonas
Dr. Yoo,

I just wanted to reiterate. I am going to create another dataset where there are going to be three columns per gene with the following names:

    GENENAME_EXP with three states (UP, DOWN, and everything else)

    GENENAME_AMP with two states (AMP;UP/DOWN; and everything else)

    GENENAME_MUT with two states (MUT/HOMDEL;UP/DOWN; and everything else)

If so, when I come back tomorrow, I'll set it up as such and run it.

Re: Datasets

PostPosted: Tue Jan 27, 2015 5:36 pm
by cwyoo
meninonas wrote:Dr. Yoo,

I just wanted to reiterate. I am going to create another dataset where there are going to be three columns per gene with the following names:

    GENENAME_EXP with three states (UP, DOWN, and everything else)

    GENENAME_AMP with two states (AMP;UP/DOWN; and everything else)

    GENENAME_MUT with two states (MUT/HOMDEL;UP/DOWN; and everything else)

If so, when I come back tomorrow, I'll set it up as such and run it.


It should be:

    GENENAME_EXP with three states (anywhere there is UP, anywhere there is DOWN, and everything else)

    GENENAME_AMP with two states (anywhere there is AMP and everything else)

    GENENAME_MUT with two states (anywhere there is MUT/HOMDEL and everything else)

Re: Datasets

PostPosted: Wed Jan 28, 2015 6:58 pm
by meninonas
Empty

Re: Datasets

PostPosted: Wed Jan 28, 2015 9:01 pm
by cwyoo
meninonas wrote:Dr. Yoo,

I have worked on the dataset below. Tomorrow, I'll continue working on the dataset tomorrow.


Always use reference coding, i.e., in GENE_EXP you have 0, 1, 2 for "anywhere there is DOWN", "all others", "anywhere there is UP" respectively; in GENE_MUT, you have 0, 1 for "all others", "anywhere there is MUT" respectively; GENE_AMP, you have 0, 1 for "all others", "anywhere there is AMP" respectively.

Create three tabs for GENE_EXP, GENE_MUT, and GENE_AMP respectively.

Re: Datasets

PostPosted: Thu Jan 29, 2015 7:33 am
by cwyoo
meninonas wrote:Dr. Yoo,

I installed Banjo on the Virtual Machine and on the computer and the following error was given to me:

Code: Select all
-----------------------------------------------------------------------------
(Final Checkpoint, after post-processing) A final check revealed the following issues that were encountered during Banjo's execution:
-----------------------------------------------------------------------------
(Post-processing) Postprocessing cannot proceed because we can't process the high score network(s).


I ran it on path4 and the following is the error that proceeds the above line:

Code: Select all
[ERROR: Banjo 2.2.0, 1/29/15 6:21:16 AM]
The following info is provided to assist in trouble-shooting:
The input line ' 89 0' in 'static.mandatory.str' is a repeat entry for variable '89'.



Stack trace info:

edu.duke.cs.banjo.bayesnet.EdgesAsArrayWithCachedStatistics.loadStructureFile(EdgesAsArrayWithCachedStatistics.java:2862)
edu.duke.cs.banjo.bayesnet.EdgesAsArrayWithCachedStatistics.<init>(EdgesAsArrayWithCachedStatistics.java:633)
edu.duke.cs.banjo.bayesnet.BayesNetManager.<init>(BayesNetManager.java:273)
edu.duke.cs.banjo.learner.SearcherSimAnneal.setupSearch(SearcherSimAnneal.java:1092)
edu.duke.cs.banjo.learner.SearcherSimAnneal.<init>(SearcherSimAnneal.java:723)
edu.duke.cs.banjo.application.BanjoThreadHandler.<init>(BanjoThreadHandler.java:115)
edu.duke.cs.banjo.application.Banjo.runMultipleInstances(Banjo.java:200)
edu.duke.cs.banjo.application.Banjo.execute(Banjo.java:150)
edu.duke.cs.banjo.application.Banjo.main(Banjo.java:447)

-----------------------------------------------------------------------------
                           End of error notification
-----------------------------------------------------------------------------
Error message while trying to record the error:
java.lang.NullPointerException


This is due to the setting that you have indicated that you are going to use:

Code: Select all
 mustBePresentEdgesFile =        static.mandatory.str


I have commented it out and it runs successfully in path4.

Code: Select all
## mustBePresentEdgesFile =        static.mandatory.str


However, we need to use above setting and the following setting in GENE_EXP, GENE_MUT, and GENE_AMP combined data:

Code: Select all
mustNotBePresentEdgesFile =


Please look into those options and learn how to specify the edges in the files.

Re: Datasets

PostPosted: Fri Feb 06, 2015 6:01 pm
by meninonas
Dr. Yoo,

Please find 10 of the 11 datasets below. They are already coded and set up.