I want to recompile BPP 3.2 code to make it update the rjMCMC state every 1% instead of every 10% on screen (analysis 11) What lines should I modify?
Aditionally, I want to set the starting tree by keyboard. I've searched extensively, in the bpp.c file, for the following lines of code specified in the bppDOC.pdf:
/ is = sptree.nspecies; printf("\nNode to collaps (a number between %d and %d, also %d)? ", sptree.nspecies+1, 2sptree.nspecies‐1, 2sptree.nspecies); scanf("%d", &is); is‐‐; /
But the the bpp.c file does not have these lines of code, instead it has this code:
/* to specify the starting tree at the keyboard, uncomment the following. /
/
printf("starting tree (1-%d)? ", sptree.nModels);
scanf("%d", &k);
k--; /
Should I remove / and / to enable setting the starting tree by keyboard?
Last edit: JCAM 2016-01-23
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Change the number 50 on the first line to 100 or a multiple of 100 such as 200 or 1000.
Change the number 20 on the second line to 100, so that you get one line on the screen for every 1% of the MCMC iteration.
use a small dataset to test.
(ii)
yes, remove the
/ /
as instructed in the code. i'll update the doc. it was not updated when the code changed.
i think this works with A10, but i am not sure what it does for A11.
If you are running bpp from the command line, you can always start the run, look at the startging tree or the number of parameters printed on the monitor, and kill the job (Ctrl-C) if you don't like the starting tree. Try a few times until you are happy with the starting tree.
best,
ziheng
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Great, although the tree-by-keyboard code does not work with A11. It can be a little frustrating to set the starting tree you want, specially when you fix the seed number.
Moving on a different other topic.
I have an issue with the original .exe, possibly a bug.
I've ran several parallel analyses before, using 4-5 cores in the aformentioned PC. Sometimes, I've used separate folders in different directories (levels) of the hard drive, and other times different folders on the same parent directory. But recently, it seems to be a major issue, I can't even run a single analysis. When I try to run the BPP 3.2 binary with my .ctl file, the command promt freezes. Meanwhile, if I check the Win7 processes manager I can see that there are not one, but three bpp.exe processes running. Strangely these processes cannot be killed, not even using the taskkill /F /IM command. I can't recall seeing these three bpp processes before.
This happens every time I attempt to run BPP 3.2, even after restarting the PC. I've tried erasing every bpp.exe on the hard drive and downloading it again but still the same outcome. I'have CCleaner installed, and if I scann the registry it brings out this issue:
"Missing MUI reference for bpp.exe"
I'm not sure what is going on and why I can't run BPP 3.2 any more. Any insight on this issue?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
i haven't seen anything like this before.
i compiled bpp.exe on a intel machine using MSC++ compiler. I wonder whether that may be an issue on your AMD. can you get gcc/cygwin and compile yourself and see whether that makes a difference.
ziheng
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Dear Dr. Z. Yang,
I want to recompile BPP 3.2 code to make it update the rjMCMC state every 1% instead of every 10% on screen (analysis 11) What lines should I modify?
Aditionally, I want to set the starting tree by keyboard. I've searched extensively, in the bpp.c file, for the following lines of code specified in the bppDOC.pdf:
/ is = sptree.nspecies; printf("\nNode to collaps (a number between %d and %d, also %d)? ", sptree.nspecies+1, 2sptree.nspecies‐1, 2sptree.nspecies); scanf("%d", &is); is‐‐; /
But the the bpp.c file does not have these lines of code, instead it has this code:
/* to specify the starting tree at the keyboard, uncomment the following. /
/
printf("starting tree (1-%d)? ", sptree.nModels);
scanf("%d", &k);
k--;
/
Should I remove / and / to enable setting the starting tree by keyboard?
Last edit: JCAM 2016-01-23
(i)
if(noisy && (k<=500 || (ir+1)%(k/50)==0)) { /* printing on screen */...
if(mcmc.sampfreq*mcmc.nsample>=50 && (ir+1)%(mcmc.sampfreq*mcmc.nsample/20)==0) {Change the number 50 on the first line to 100 or a multiple of 100 such as 200 or 1000.
Change the number 20 on the second line to 100, so that you get one line on the screen for every 1% of the MCMC iteration.
use a small dataset to test.
(ii)
yes, remove the
/
/
as instructed in the code. i'll update the doc. it was not updated when the code changed.
i think this works with A10, but i am not sure what it does for A11.
If you are running bpp from the command line, you can always start the run, look at the startging tree or the number of parameters printed on the monitor, and kill the job (Ctrl-C) if you don't like the starting tree. Try a few times until you are happy with the starting tree.
best,
ziheng
Great, although the tree-by-keyboard code does not work with A11. It can be a little frustrating to set the starting tree you want, specially when you fix the seed number.
Moving on a different other topic.
I have an issue with the original .exe, possibly a bug.
My PC specs are as follows:
OS: Windows 7 Ultimate.
CPU: AMD fx6300 six-core @ 3.5 Ghz each.
I've ran several parallel analyses before, using 4-5 cores in the aformentioned PC. Sometimes, I've used separate folders in different directories (levels) of the hard drive, and other times different folders on the same parent directory. But recently, it seems to be a major issue, I can't even run a single analysis. When I try to run the BPP 3.2 binary with my .ctl file, the command promt freezes. Meanwhile, if I check the Win7 processes manager I can see that there are not one, but three bpp.exe processes running. Strangely these processes cannot be killed, not even using the taskkill /F /IM command. I can't recall seeing these three bpp processes before.
This happens every time I attempt to run BPP 3.2, even after restarting the PC. I've tried erasing every bpp.exe on the hard drive and downloading it again but still the same outcome. I'have CCleaner installed, and if I scann the registry it brings out this issue:
"Missing MUI reference for bpp.exe"
I'm not sure what is going on and why I can't run BPP 3.2 any more. Any insight on this issue?
i haven't seen anything like this before.
i compiled bpp.exe on a intel machine using MSC++ compiler. I wonder whether that may be an issue on your AMD. can you get gcc/cygwin and compile yourself and see whether that makes a difference.
ziheng
View and moderate all "BP&P Q&A" comments posted by this user
Mark all as spam, and block user from posting to "Discussion"
great thread