Investigate why Travis tests failing

[pgleeson]

@andrisecker can you look here: https://travis-ci.org/pgleeson/ca1/jobs/272150687

The NEURON omv tests are failing, with similar errors to what I get locally. Some differences are small, but others, e.g. olm are larger:

[omv] Running tests defined in .test.olm.nrn.omt
[omv] ==========================================
[omv]   Checking whether OMVEngine: NEURON is already installed...
[omv]     Running the commands: ['nrniv', '--version'] in (.; cwd=/home/padraig/git/ca1/NeuroML2; shell=False)
[omv]     Commands: ['nrniv', '--version'] completed successfully
[omv]     NEURON -- Release 7.4 (1370:16a7055d4a86) 2015-11-09 was already installed locally
[omv]   Checking whether NEURON is already installed...
[omv]     Running the commands: ['nrniv', '--version'] in (.; cwd=/home/padraig/git/ca1/NeuroML2; shell=False)
[omv]     Commands: ['nrniv', '--version'] completed successfully
[omv]     NEURON -- Release 7.4 (1370:16a7055d4a86) 2015-11-09 was already installed locally
[omv]   NEURON environment vars: {'PYTHONPATH': '/home/padraig/local/lib/python/site-packages', 'NEURON_HOME': '/usr/local/nrn/x86_64/'}, PATH: /usr/local/nrn/x86_64/bin
[omv]     Set env var PYTHONPATH: /home/padraig/opt/nest/lib64/python2.7/site-packages:/home/padraig/opt/nest/lib/python2.7/site-packages::/home/padraig/local/lib/python/site-packages
[omv]     Set env var NEURON_HOME: /usr/local/nrn/x86_64/
[omv]     Setting PATH: /home/padraig/opt/nest/bin:/home/padraig/bin:/usr/local/nrn/x86_64//bin:/usr/local/genesis:/home/padraig/moose:/sbin/:/home/padraig/neuroConstruct/NeuroML2:/home/padraig/jLEMS/:/home/padraig/neuroConstruct/jNeuroMLJar:/home/padraig/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin:/usr/lib/jvm/java-8-oracle/bin:/usr/lib/jvm/java-8-oracle/db/bin:/usr/lib/jvm/java-8-oracle/jre/bin:/usr/local/cuda/bin:/usr/local/nrn/x86_64/bin
[omv]     Attention! Preexistent datafile: olm.soma.dat
[omv]   Found 1 experiment(s) to run on engine: NEURON 
[omv]   Running NEURON on /home/padraig/git/ca1/NeuroML2/olm.hoc...
[omv]     OUT: 	1 
##### olm started #####
Starting simulation of duration 300 ms, dt: 0.001 ms
Finished simulation in 2.3199999 seconds

No plots shown. Change variable show_plot to view membrane potential traces

	1 

[omv]     ERR: NEURON -- Release 7.4 (1370:16a7055d4a86) 2015-11-09
Duke, Yale, and the BlueBrain Project -- Copyright 1984-2015
See http://www.neuron.yale.edu/neuron/credits

Additional mechanisms from files
 ../ExpGABAab.mod ../MyExp2Sid.mod ../MyExp2Sidnw.mod ../SIN.mod ../ch_CavL.mod ../ch_CavN.mod ../ch_HCN.mod ../ch_HCNolm.mod ../ch_HCNp.mod ../ch_KCaS.mod ../ch_Kdrfast.mod ../ch_Kdrfastngf.mod ../ch_Kdrp.mod ../ch_Kdrslow.mod ../ch_KvA.mod ../ch_KvAdistp.mod ../ch_KvAngf.mod ../ch_KvAolm.mod ../ch_KvAproxp.mod ../ch_KvCaB.mod ../ch_KvGroup.mod ../ch_KvM.mod ../ch_Nav.mod ../ch_Navaxonp.mod ../ch_Navbis.mod ../ch_Navcck.mod ../ch_Navngf.mod ../ch_Navp.mod ../ch_leak.mod ../fastconn.mod ../iconc_Ca.mod ../mynetstim.mod ../positionfcns.mod ../repeatconn.mod ../sgate.mod ../xtra.mod
nrniv: unable to open font "*helvetica-medium-r-normal*--14*", using "fixed"

[omv]     returncode: 
[omv]   Running checks for experiment: current clamp olm
[omv] 
[omv]   Reading timeseries from: {'initial rows to disregard': '0', 'file name': 'olm.soma.dat', 'columns(time, voltage)': '[0, 1]'}
[omv]     Detecting spikes with method: threshold
[omv] Comparison of 
                    (observed data): [64.044, 81.16, 96.782, 111.747, 126.376, 140.823, 155.17, 169.464, 183.734, 197.997, 212.263, 226.538, 240.827]
                    and
                    (expected data): [62.433, 78.643, 93.897, 108.661, 123.153, 137.488, 151.73, 165.918, 180.075, 194.215, 208.349, 222.48, 236.614, 250.856]
                    failed against tolerance 0
[omv]       Observable                        Test Passed     
[omv]       --------------------------------------------------
[omv]       spike times                            ✘          
[omv]       +++++++++++++++++++++ Error info ++++++++++++++++++
[omv]        Return code: 0
[omv]        Output: 	1 
       ##### olm started #####
       Starting simulation of duration 300 ms, dt: 0.001 ms
       Finished simulation in 2.3199999 seconds
       
       No plots shown. Change variable show_plot to view membrane potential traces
       
       	1 
       
[omv]       +++++++++++++++++++++++++++++++++++++++++++++++++++
[omv] 
[omv]                     Test failed: .test.olm.nrn.omt   

[andrisecker]

@pgleeson I have no clue what's happening... The NeuroML2 ones are still passing with the same .mep file and we didn't touch the .mod files... (or anything in those models)

Might be the same as issue #13
They're also failing for me locally with version 7.4 1370 (same as on travis). Just asked the NEURON developers about it: https://www.neuron.yale.edu/phpBB/viewtopic.php?f=6&t=3747

-> Update on this: NEURON version 7.5 didn't solve the issue (but crashed the .mod file generated from Capool.nml) so we stick to 7.4

[andrisecker]

To be honest we skipped checking through the changes since our first clone from bitbucket and that's why some travis tests were failing...

• CCK+BC was failing since big changes were made in Marianne's last bitbucket commit. The conductances are now set to the values reported in eLife paper as in bitbucket version f1bb857, however 2 channels: Navcck and KCaS needed to be as they were in version f1bb857, since the ones in version _ 0e789a2 _ and also reported in the paper don't lead to the behaviour shown in http://mariannebezaire.com/ca1_graphic/celltype.html#cckcell and in the paper
  -> this 2 channel descriptions are inconsistent with the results in the eLife paper
  I've created new versions of the channels and the reoptimized cells and both versions are passing now

• Cutsuridis cell (as a simpler PC template) is not even mentioned in the paper so I couldn't check which version is the correct... I've just setted it back to bitbucket version f1bb857 since that was passing.

• Ivy and NGF cells were failing since the update in KCaS, which I just set back to the previous one as in the case of CCK (and SCA)

• OLM cell was failing since h ion wasn't inserted (see issue #16) - that's fixed and with conductances matching the eLife paper see bitbucket version f1bb857 it's passing again

• SCA cell is the same as CCK+BC

@pgleeson Everything is passing for me!

[pgleeson]

Only one h5 based Netpyne network simulation is failing on Travis (and my machine) now... https://travis-ci.org/pgleeson/ca1/jobs/299699916