Skip to content

Add worker that writes mismatch data to disk file#7

Merged
timholy merged 7 commits into
masterfrom
teh/Mat
Oct 22, 2015
Merged

Add worker that writes mismatch data to disk file#7
timholy merged 7 commits into
masterfrom
teh/Mat

Conversation

@timholy
Copy link
Copy Markdown
Member

@timholy timholy commented Oct 17, 2015

This prepares the way for testing an older strategy---explicitly finding the minimum over an entire experiment, not just on a stack-by-stack basis---with the new registration code. The key commit is 2698f2b. It's mildly outrageous how easy this was to add, a possible sign that the Driver/Worker architecture is well-designed 😄.

Still working on actually using this information for optimization (that is much harder).

While this PR will fail due to the julia bug in the module paths, locally (on cannon) it should now be possible to run runtests.jl.

This is needed if we monitor Qs from RegisterFit and store it as an Array{Mat}.
This prepares the way for including a temporal penalty on the deformation to enforce gradual change.
When running all tests through runtests, sometimes the workers hadn't been killed when a new `using...` statement in the next test ran. This could lead to path errors due to julia#3674.
@timholy timholy mentioned this pull request Oct 17, 2015
@Cody-G
Copy link
Copy Markdown

Cody-G commented Oct 17, 2015

Wow, I think this could be a big breakthrough!!

@timholy
Copy link
Copy Markdown
Member Author

timholy commented Oct 17, 2015

It also could be unusably slow, if we want to register each stack. The mismatch computation is fast, it's the optimization I'm worried about---optimizing 10^7 parameters simultaneously is not child's play. We'll see how it works out.

This is needed for reasonable performance, and especially mmapping, when writing mismatch data to disk.
timholy added a commit that referenced this pull request Oct 22, 2015
Add worker that writes mismatch data to disk file
@timholy timholy merged commit 832b5df into master Oct 22, 2015
@timholy timholy deleted the teh/Mat branch October 22, 2015 12:49
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants