search

zxzhijia / Brian2STDPMNIST

Brian 2 version of Paper "Unsupervised Learning of digit recognition using STDP"
128 stars 71 forks source link

Old Brian 2 port #4

Closed mrahtz closed 8 years ago

mrahtz commented 8 years ago

This is my old port of Brian 2. I haven't tested the classification accuracy, but it seems to be producing reasonable spike output. (It's not exactly the same output as the original Brian 1 version, but I'm guessing that's because of a difference in PoissonGroup implementation between Brian 1 and 2 which means that the input is slightly different.)

zxzhijia commented 8 years ago

Thank you very much. I've merged it.

On Tuesday, September 13, 2016, Matthew Rahtz notifications@github.com wrote:

This is my old port of Brian 2. I haven't tested the classification accuracy, but it seems to be producing reasonable spike output. (It's not exactly the same output as the original Brian 1 version, but I'm guessing that's because of a difference in PoissonGroup implementation between

Brian 1 and 2 which means that the input is slightly different.)

You can view, comment on, or merge this pull request online at:

https://github.com/zxzhijia/Brian2STDPMNIST/pull/4 https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_zxzhijia_Brian2STDPMNIST_pull_4&d=CwMCaQ&c=imBPVzF25OnBgGmVOlcsiEgHoG1i6YHLR0Sj_gZ4adc&r=X3YsXL1AZsKHgsh28HFY8A&m=tqMAtgClMAOntsgeG5XXkJnwtPTWhJPNgr7-d0oH9EM&s=Qaw_FnZd5Mb7idW_SUSdVMN8qygCNv1SufPTWP-0tp8&e= Commit Summary

  • Remove trailing whitespace
  • s/b/b2/g
  • Remove old args to NeuronGroup
  • Fix unit definition
  • Fix threshold/reset
  • Fix subgroup calculation
  • Fix connections
  • Don't load Brian 1
  • Fix rate monitors
  • Fix input rate
  • Rename a couple more b. to b2.
  • Fix another PopulationRateMonitor
  • Start fixing second set of Connections
  • continue fixing second connect
  • Fix connections
  • Fix STDP
  • Fix connection plotting
  • Fix rates setting
  • Fix initial run command
  • Hopefully fix weight normalisation
  • Fix rates units
  • Fix network setup
  • Fix connection saving
  • Fix refrac string
  • Fix problem with delays
  • Remove global preferences
  • Fix theta import
  • Fix deprecated pre/post syntax
  • Remove real-time spike plotting
  • Fix warnings about integration method
  • Fix warning about rates
  • Hopefully slightly speed up simulation
  • Fix subplot plot numbering
  • REALLY fix theta loading
  • Fix refractoriness
  • Fix monitors
  • Remove extra spike monitor
  • Fix refractoriness properly

File Changes

Patch Links:

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_zxzhijia_Brian2STDPMNIST_pull_4&d=CwMCaQ&c=imBPVzF25OnBgGmVOlcsiEgHoG1i6YHLR0Sj_gZ4adc&r=X3YsXL1AZsKHgsh28HFY8A&m=tqMAtgClMAOntsgeG5XXkJnwtPTWhJPNgr7-d0oH9EM&s=Qaw_FnZd5Mb7idW_SUSdVMN8qygCNv1SufPTWP-0tp8&e=, or mute the thread https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_ALPsQNtWzM-5FvaXBohEM8UyixoqboujY1ks5qpnQ1gaJpZM4J7fts&d=CwMCaQ&c=imBPVzF25OnBgGmVOlcsiEgHoG1i6YHLR0Sj_gZ4adc&r=X3YsXL1AZsKHgsh28HFY8A&m=tqMAtgClMAOntsgeG5XXkJnwtPTWhJPNgr7-d0oH9EM&s=Linc8A9BABS-ur7xFiVAc-03pzwPv1xwSTlNEGwy9kg&e= .