def get_model(num_timesteps, num_input_vars, num_output_vars, p_params=None):
params = dict(archtype='rnn',
optimizer='rmsprop',
discount=0.9,
hidden_layer_dims=[])
if p_params:
params.update(p_params)
#params.update( dict(macro_dims=500, archtype='RNN',hidden_activation='leakyrelu') )
#params.update( dict(macro_dims=500, hidden_activation='leakyrelu') ) # gets better than mean perfornace on ieee300
#params.update( dict(macro_dims=trajs_trn.shape[2], archtype='RNNIdentity', hidden_activation='leakyrelu') ) # gets better than mean perfornace on ieee300
#params.update( dict(macro_dims=trajs_trn.shape[2], archtype='RNNIdentity', hidden_activation='linear') ) # gets better than mean perfornace on ieee300
#params.update( dict(macro_dims=500, hidden_activation='srelu') )# , hidden_layer_dims=[500,])
model_name = "models/" + "_".join(["%s"%x[1] for x in sorted(params.items()) if x[1]]) + \
"-" + "-".join(map(str, [num_timesteps, num_input_vars, num_output_vars]))
if False:
# elsewhere...
from keras.models import model_from_json
with timeIt("Loading model object"):
model = model_from_json(open('%s.json' % model_name).read())
model.load_weights('%s_weights.h5' % model_name)
else:
with timeIt("Creating model object"):
model = trainrnn.get_rnn_model(num_timesteps,
num_input_vars,
num_output_vars=num_output_vars,
output_type='real',
**params)
model.model_name = model_name
return model
def get_model(num_timesteps, num_input_vars, num_output_vars, p_params=None):
params = dict(archtype='rnn',optimizer='rmsprop',discount=0.9, hidden_layer_dims=[])
if p_params:
params.update( p_params )
#params.update( dict(macro_dims=500, archtype='RNN',hidden_activation='leakyrelu') )
#params.update( dict(macro_dims=500, hidden_activation='leakyrelu') ) # gets better than mean perfornace on ieee300
#params.update( dict(macro_dims=trajs_trn.shape[2], archtype='RNNIdentity', hidden_activation='leakyrelu') ) # gets better than mean perfornace on ieee300
#params.update( dict(macro_dims=trajs_trn.shape[2], archtype='RNNIdentity', hidden_activation='linear') ) # gets better than mean perfornace on ieee300
#params.update( dict(macro_dims=500, hidden_activation='srelu') )# , hidden_layer_dims=[500,])
model_name = "models/" + "_".join(["%s"%x[1] for x in sorted(params.items()) if x[1]]) + \
"-" + "-".join(map(str, [num_timesteps, num_input_vars, num_output_vars]))
if False:
# elsewhere...
from keras.models import model_from_json
with timeIt("Loading model object"):
model = model_from_json(open('%s.json'%model_name).read())
model.load_weights('%s_weights.h5' % model_name)
else:
with timeIt("Creating model object"):
model = trainrnn.get_rnn_model(num_timesteps, num_input_vars, num_output_vars=num_output_vars, output_type='real', **params)
model.model_name = model_name
return model
if t == 0:
if clocs is not None:
if len(clocs) == 15:
yield np.vstack(df.Effs2.iloc[clocs])
clocs = []
clocs.append(ndx)
if len(clocs) == 15:
yield np.vstack(df.Effs2.iloc[clocs])
trajs = np.stack(split_mx(df))
import trainrnn
import logging
logging.getLogger('keras').setLevel(logging.INFO)
hidden_dims = 350
model = trainrnn.get_rnn_model(dshape=trajs.shape,
hidden_dims=hidden_dims,
output_type='real')
# model.fit(trajs, trajs, nb_epoch=100, batch_size=10, validation_split=0.1, show_accuracy=True, verbose=1)
#model.fit(trajs, trajs, nb_epoch=10, batch_size=5, validation_split=0.1, show_accuracy=True, verbose=1)
model.fit(trajs,
trajs,
nb_epoch=100,
batch_size=1,
validation_split=0.1,
show_accuracy=True,
verbose=1)
import numpy as np
df = pd.read_pickle('out1.pkl')
def split_mx(df):
clocs = None
for ndx, t in enumerate(df[df.PertId!=-1].t):
if t == 0:
if clocs is not None:
if len(clocs) == 15:
yield np.vstack(df.Effs2.iloc[clocs])
clocs = []
clocs.append(ndx)
if len(clocs) == 15:
yield np.vstack(df.Effs2.iloc[clocs])
trajs = np.stack(split_mx(df))
import trainrnn
import logging
logging.getLogger('keras').setLevel(logging.INFO)
hidden_dims = 350
model = trainrnn.get_rnn_model(dshape=trajs.shape, hidden_dims=hidden_dims, output_type='real')
# model.fit(trajs, trajs, nb_epoch=100, batch_size=10, validation_split=0.1, show_accuracy=True, verbose=1)
#model.fit(trajs, trajs, nb_epoch=10, batch_size=5, validation_split=0.1, show_accuracy=True, verbose=1)
model.fit(trajs, trajs, nb_epoch=100, batch_size=1, validation_split=0.1, show_accuracy=True, verbose=1)
