realloc() error

[biswajitSM]

Hi @nsong06, I compiled the changepoint code and ran the following command for processing: mpirun -np 2 Parallel-GM file.dat 0.05 0.95 But it showed error after checking for the nodes. Here is how the error looks like:

Have you faced this problem ever. Can you please suggest a solution for this. System info: Intel® Core™ i7-6500U CPU @ 2.50GHz × 4 , 64 bit ubuntu 16.04 LTS GNU C Library 2.23 OpenMPI 1.10.2

Thanks Biswajit

[nsong06]

Hi Biswajit,

I have never seen the error before; however, after looking over the code, I believe I've found the error in main.c and fixed it on the master branch. Try running the new code and let me know if there are any more issues.

Thanks Nancy

[biswajitSM]

Thanks a lot @nsong06 . Now it works without error and quite fast. I compared with the results from changepoint-2005 program. And it gives way many changepoint with the same confidence interval. The Parallel-GM gives 10285 change points while the 2005-program gives 1385 points. The output from 2005-program fits well with the real binned trace (visually) while the output from parallel-GM doesn't.

Am I doing something wrong? Here is the the file with photon arrival time. Point_0.11uW_A5_0mV(66)_1.dat.tar.gz

the command was mpirun -np 2 Parallel-GM 'Point_0.11uW_A5_0mV(66)_1.dat' 0.1 0.9

And the last few lines of command out put was:

I have forked the repository and will add more python-usage (analysis with examples) once it works properly.

Again thanks a lot for your time Biswajit

[nsong06]

Hi Biswajit, This program uses a different analysis and data than the 2005 program - the 2005 program takes photon arrival time as input, but this program requires the fluorescence intensity as input. You will note that the 2005 program uses a Poisson distribution to model photon count while this uses a Gaussian distribution.

Nancy

[biswajitSM]

Ah I see. stupid me. Anyway it's useful for many general purpose. thanks a lot!