entry = pairs.get(dataset.LABELS[i])
if entry is None:
continue
if len(entry['x']) != 2:
continue
color = COLOR_MAP(to_color(i))
plt.plot(entry['x'], entry['y'], color=color)
if fname == None:
plt.show()
else:
plt.savefig(fname=fname)
print("Saved image to " + fname)
if __name__ == '__main__':
import sys
datasets = dataset.parse()
siamese, _, _ = gradtype_model.create()
siamese.load_weights(sys.argv[1])
train_datasets, validate_datasets = dataset.split(datasets)
train_datasets = dataset.trim_dataset(train_datasets)
validate_datasets = dataset.trim_dataset(validate_datasets)
train_coords = dataset.evaluate_model(siamese, train_datasets)
validate_coords = dataset.evaluate_model(siamese, validate_datasets)
fname = sys.argv[2] if len(sys.argv) >= 3 else None
pca(train_coords, validate_coords, fname)
saver = tf.train.Saver(max_to_keep=0, name='hist-dist')
restore = sys.argv[1]
if restore.endswith('.index'):
restore = restore[:-6]
saver.restore(sess, restore)
logging.debug('Loading dataset...')
loaded = dataset.load(overlap=8)
train_dataset = loaded['train']
validate_dataset = loaded['validate']
logging.debug('Trimming dataset...')
train_dataset, _ = dataset.trim_dataset(train_dataset, random_state=SEED)
validate_dataset, _ = dataset.trim_dataset(validate_dataset,
random_state=SEED)
train_dataset, _ = dataset.flatten_dataset(train_dataset,
random_state=SEED)
validate_dataset, _ = dataset.flatten_dataset(validate_dataset,
random_state=SEED)
holds = []
codes = []
deltas = []
sequence_lens = []
for seq in train_dataset:
holds.append(seq['holds'])
model.compile(adam,
loss='categorical_crossentropy',
metrics=['accuracy', top_5])
#
# Train
#
tb = TensorBoard(histogram_freq=50,
log_dir=gradtype_utils.get_tensorboard_logdir())
callbacks = [tb]
for i in range(start_epoch, TOTAL_EPOCHS, SAVE_EPOCHS):
end_epoch = i + SAVE_EPOCHS
train_x = dataset.gen_regression(dataset.trim_dataset(train_datasets))
train_y = gradtype_model.generate_one_hot_regression(train_x['labels'])
model.fit(x=train_x,
y=train_y,
batch_size=4096,
initial_epoch=i,
epochs=end_epoch,
callbacks=callbacks,
validation_data=(validate_x, validate_y))
print("Saving...")
fname = './out/gradtype-regr-{:08d}.h5'.format(end_epoch)
siamese.save_weights(fname)