def loadExperiment(self, experiment):
suite = Suite()
suite.parse_opt()
suite.parse_cfg()
experiment_dir = experiment.split('/')[1]
params = suite.items_to_params(suite.cfgparser.items(experiment_dir))
self.params = params
predictions = suite.get_history(experiment, 0, 'predictions')
truth = suite.get_history(experiment, 0, 'truth')
self.iteration = suite.get_history(experiment, 0, 'iteration')
self.train = suite.get_history(experiment, 0, 'train')
self.truth = np.array(truth, dtype=np.float)
if params['output_encoding'] == 'likelihood':
from nupic.encoders.scalar import ScalarEncoder as NupicScalarEncoder
self.outputEncoder = NupicScalarEncoder(w=1, minval=0, maxval=40000, n=22, forced=True)
predictions_np = np.zeros((len(predictions), self.outputEncoder.n))
for i in xrange(len(predictions)):
if predictions[i] is not None:
predictions_np[i, :] = np.array(predictions[i])
self.predictions = predictions_np
else:
self.predictions = np.array(predictions, dtype=np.float)
predData_TM_n_step = np.roll(np.array(predict_data_ML),
_options.stepsAhead)
nTest = len(actual_data) - nTrain - _options.stepsAhead
NRMSE_TM = NRMSE(actual_data[nTrain:nTrain + nTest],
predData_TM_n_step[nTrain:nTrain + nTest])
print "NRMSE on test data: ", NRMSE_TM
output.close()
# calculate neg-likelihood
predictions = np.transpose(likelihoodsVecAll)
truth = np.roll(actual_data, -5)
from nupic.encoders.scalar import ScalarEncoder as NupicScalarEncoder
encoder = NupicScalarEncoder(w=1,
minval=0,
maxval=40000,
n=22,
forced=True)
from plot import computeLikelihood, plotAccuracy
bucketIndex2 = []
negLL = []
minProb = 0.0001
for i in xrange(len(truth)):
bucketIndex2.append(np.where(encoder.encode(truth[i]))[0])
outOfBucketProb = 1 - sum(predictions[i, :])
prob = predictions[i, bucketIndex2[i]]
if prob == 0:
prob = outOfBucketProb
if prob < minProb:
prob = minProb
def __init__(self):
self.encoder = NupicScalarEncoder(w=1,
minval=0,
maxval=40000,
n=22,
forced=True)