bcgsc / physlr

:chains: Construct a Physical Map from Linked Reads
GNU General Public License v3.0
18 stars 8 forks source link

Make subgraph c++ #155

Closed aafshinfard closed 4 years ago

aafshinfard commented 4 years ago

I ran this on hg004...m92.5.tsv with strategy being bc and it took 19 mins to complete: Elapsed (wall clock) time (h:mm:ss or m:ss): 19:38.58 and it used 18 GB of ram. With the current code it runs in almost 26 mins and with 14 GB of ram.

Next is to make it parallel and also add binning to it. Then I will merge my other branches to add to this other community detection algorithms.

jwcodee commented 4 years ago

I ran this on hg004...m92.5.tsv with strategy being bc and it took 19 mins to complete: Elapsed (wall clock) time (h:mm:ss or m:ss): 19:38.58 and it used 18 GB of ram. With the current code it runs in almost 26 mins and with 14 GB of ram.

Next is to make it parallel and also add binning to it. Then I will merge my other branches to add to this other community detection algorithms.

I dont understand. What took 19 mins to complete? Python bc or the code before your changes?

aafshinfard commented 4 years ago

With the current 'master', it takes around 26 minutes (14 GB ram). With this branch we are preparing to push, it takes around 19 minutes (18 GB ram) and now we can multi-thread (I am doing it along with binning the subgraphs in another branch) .

aafshinfard commented 4 years ago

I just tested the last version of this code, and the tsv file has the same number of lines as the one I got from master. Also tested on the test data and it worked.

jwcodee commented 4 years ago

With the current 'master', it takes around 26 minutes (14 GB ram). With this branch we are preparing to push, it takes around 19 minutes (18 GB ram) and now we can multi-thread (I am doing it along with binning the subgraphs in another branch) .

Very good. Thanks for doing this!