These are chat archives for thunder-project/thunder

20th
Apr 2015
Richard A Hofer
@rhofour
Apr 20 2015 03:07
This message was deleted
tomsains
@tomsains
Apr 20 2015 10:27

@freeman-lab (or anyone else who knows) i'm having a little bit of difficulty implementing the circular tuning model. I have regressed to all on my stimulus directions (12) and now want to fit a circular tuning model to the betas so that i can estimate direction selectivity. What i am unclear on is what 'stimparams' are needed in the following line of code:

tuningmodel = TuningModel.load(modelfile='stimparams', tuningmode='circular')

tomsains
@tomsains
Apr 20 2015 14:17

@freeman-lab
I have just used thunder on EC2, but for some reason I am having trouble loading data from S3 via the following method:

images = tsc.loadImages('s3n://light-sheet-data/24-2-15_gc6_4dpf_BTX_20Hz_12pl_30_20_8um_20min_no.1.2_mt/', inputFormat='tif')

but the this error is being returned:

S3ResponseError: S3ResponseError: 301 Moved Permanently

I have not used thunder on a cluster in a couple of weeks and i have just upgraded to spark1.3.0, are there any changes that have been made which might explain this? do i now have to configure my AWS credentials?

I can however load the exampleEC2 datasets without any issues.

Richard A Hofer
@rhofour
Apr 20 2015 23:28
Inverting matrices is hard :(