from spectrogram import full_bpm_to_data, HEART_AV_ROOT, NormalizedSpectrograms
from get_heartrates import get_interesting_heartrates
from keras.callbacks import EarlyStopping
from kbhit import KBHit
import numpy as np
import code
import random
import learnLib
kb = KBHit()
#(X_train, y_train), (X_test, y_test) = full_bpm_to_data(get_interesting_heartrates(HEART_AV_ROOT))
ns = NormalizedSpectrograms()
(X_train, Y_train) , valTuple = ns.getTrainAndValidationData()
print(X_train.shape)
print("Model: nb_hiddens, drop1s")
prevLoss = 34534645735673
maxModel = None
stop = False
models = {}
X_validate, Y_validate = valTuple
for args in learnLib.RandomMlpParameters(): #itertools.product(nb_hiddens, drop1s):
print("Model: ", args)
model = learnLib.get_2_layer_MLP_model(X_train[0].shape, *args)
kb = KBHit()
# (X_train, y_train), (X_test, y_test) = full_bpm_to_data(get_interesting_heartrates(HEART_AV_ROOT))
ns = NormalizedSpectrograms()
def sliceToTimeSeries(X):
divisibleTime = X[:, 0, :, :150]
slicedTime = np.reshape(divisibleTime, (-1, X.shape[2], 30, 5))
swappedAxes = np.swapaxes(slicedTime, 1, 2)
flattenLastTwo = np.reshape(swappedAxes, (X.shape[0], 30, -1))
return flattenLastTwo
(X_train, Y_train), (X_val, Y_val) = ns.getTrainAndValidationData()
# slice the spectrogram
X_train = sliceToTimeSeries(X_train)
print(X_train.shape)
# Y_train = np.repeat(np.reshape(-1,1), X_train.shape[1], axis=1)
print(Y_train.shape)
print("Model: lstm outdim, nb_hiddens, drop1, drop2")
prevLoss = 34534645735673
maxModel = None
stop = False
models = {}
X_val = sliceToTimeSeries(X_val)
