prnn.utils package
Submodules
prnn.utils.ActionEncodings module
prnn.utils.Architectures module
prnn.utils.CANNNet module
prnn.utils.CANNtools module
Created on Tue May 10 12:49:31 2022
@author: dl2820
- prnn.utils.CANNtools.CANNmatrix(Ncells, size, selfconnect=False, peak=1, width=1, inh=0)
Size: N-D list
- prnn.utils.CANNtools.expKernel(dist, width=1)
- prnn.utils.CANNtools.multiCANNmatrix(Ncells, size, Nmaps, selfconnect=False, peak=1, width=1, inh=0)
- prnn.utils.CANNtools.periodicDist(x, size)
prnn.utils.LayerNormRNN module
prnn.utils.LinearDecoder module
prnn.utils.Shell module
prnn.utils.agent module
prnn.utils.data module
prnn.utils.eg_utils module
prnn.utils.env module
prnn.utils.figures module
Created on Tue Jan 18 16:15:54 2022
@author: dl2820
- prnn.utils.figures.IsoMapFigure(predictiveNet, env, agent, noisemag=0, noisestd=0.25, timesteps_wake=5000, timesteps_sleep=1000, savename=None, savefolder=None, usecells=None)
- prnn.utils.figures.OfflineTrajectoryProps(predictiveNet, decoder, timesteps=5000, noisemag=0, logstdrange=(-2, 0.5), numpoints=21, savename=None, savefolder=None)
- prnn.utils.figures.SpontActivityExamples(predictiveNet, examples=((0, 0.1), (0, 0.25), (0, 0.4)), timesteps=4000, savename=None, savefolder=None, decoder=None)
- prnn.utils.figures.SpontActivityFigure(predictiveNet, compareWAKEagent=None, murange=0.5, maxstd=1, numpoints=11, timesteps=1000, examples=((0, 0.1), (0, 0.25), (0, 0.4)), savename=None, savefolder=None, decoder=None)
- prnn.utils.figures.SpontTrajectoryFigure(predictiveNet, decoder, noisemag=0, noisestd=0.25, timesteps=5000, savename=None, savefolder=None)
- prnn.utils.figures.TrainingFigure(predictiveNet, savename=None, savefolder=None, incDecode=True)
prnn.utils.general module
Created on Mon Dec 13 16:17:09 2021
@author: dl2820
- prnn.utils.general.clumpyRandom(size, choices, seedprobability, numiter=1)
- prnn.utils.general.delaydist(signal, numdelays=10, maxdist=15, firstdelay=1, sqdist=False, dist='cityblock')
- prnn.utils.general.fit_exp_linear(t, y, C=0)
- prnn.utils.general.kl_divergence(p, q)
- prnn.utils.general.loadPkl(savename, savepath=None)
- prnn.utils.general.mkdir_p(mypath)
Creates a directory. equivalent to using mkdir -p on the command line If the path exists, no error
- prnn.utils.general.saveFig(fig, savename, savepath=None, filetype='png', dpi=300)
- prnn.utils.general.savePkl(obj, savename, savepath=None)
- prnn.utils.general.state2nap(state)
prnn.utils.lossFuns module
prnn.utils.plotUtils module
- prnn.utils.plotUtils.setNiceAxes()
- prnn.utils.plotUtils.setPlotDefaults()