Functions — github.com/benanne/kaggle-galaxies

CodeHub MCPcopy Create free account

hub / github.com/benanne/kaggle-galaxies / functions

Functions336 in github.com/benanne/kaggle-galaxies

⨍ Functions 336◇ Types & classes 53

↓ 1 callersFunctionpost_augment_chunk

perform fast augmentation that can be applied directly to the chunks in realtime.

load_data.py:477

↓ 1 callersFunctionpost_augment_chunks

post augmentation MODIFIES THE CHUNKS IN CHUNK_LIST IN PLACE to save memory where possible! post_augmentation_params: - gamma_range:

realtime_augmentation.py:349

↓ 1 callersMethodpredictions

(self, *args, **kwargs)

↓ 1 callersMethodpredictions

(self, *args, **kwargs)

↓ 1 callersMethodpredictions

(self, *args, **kwargs)

↓ 1 callersMethodpredictions

(self, *args, **kwargs)

↓ 1 callersMethodpredictions

(self, *args, **kwargs)

↓ 1 callersMethodpredictions

(self, normalisation=True, *args, **kwargs)

↓ 1 callersMethodpredictions

(self, normalisation=True, *args, **kwargs)

↓ 1 callersMethodpredictions

(self, normalisation=True, *args, **kwargs)

↓ 1 callersMethodpredictions_no_normalisation

Predict without normalisation. This can be used for the first few chunks to find good parameters.

↓ 1 callersMethodpredictions_no_normalisation

Predict without normalisation. This can be used for the first few chunks to find good parameters.

↓ 1 callersMethodpredictions_no_normalisation

Predict without normalisation. This can be used for the first few chunks to find good parameters.

↓ 1 callersMethodrescaled_weights

↓ 1 callersMethodrescaled_weights

↓ 1 callersMethodrescaled_weights

↓ 1 callersMethodreset_params

cc_layers.py:66

↓ 1 callersMethodreset_params

cc_layers.py:291

↓ 1 callersMethodreset_params

↓ 1 callersMethodreset_params

↓ 1 callersMethodreset_params

↓ 1 callersMethodreset_params

↓ 1 callersMethodreset_params

↓ 1 callersFunctionselect_indices

(num, num_selected)

realtime_augmentation.py:40

↓ 1 callersMethodtargets

(self, *args, **kwargs)

↓ 1 callersMethodtargets

(self, *args, **kwargs)

↓ 1 callersFunctiontc_exp

A version of the exponential that returns 0 below a certain threshold.

↓ 1 callersMethodweighted_answer_probabilities

(self, weight_with_targets=False, *args, **kwargs)

↓ 1 callersMethodweighted_answer_probabilities

(self, *args, **kwargs)

↓ 1 callersMethodweighted_answer_probabilities

(self, *args, **kwargs)

↓ 1 callersMethodweighted_answer_probabilities

(self, *args, **kwargs)

↓ 1 callersMethodweighted_answer_probabilities

(self, weight_with_targets=False, *args, **kwargs)

↓ 1 callersMethodweighted_answer_probabilities

(self, *args, **kwargs)

↓ 1 callersMethodweighted_answer_probabilities

(self, *args, **kwargs)

↓ 1 callersMethodweighted_answer_probabilities

(self, *args, **kwargs)

(self, img_index)

realtime_augmentation.py:213

(self, img_index)

realtime_augmentation.py:309

(self, img_index)

realtime_augmentation.py:563

(self, img_index)

realtime_augmentation.py:587

(self, img_index)

realtime_augmentation.py:611

(self, img_index)

realtime_augmentation.py:655

(self, img_index)

realtime_augmentation.py:683

(self, img_index)

realtime_augmentation.py:711

(self, img_index)

realtime_augmentation.py:735

(self, img_index)

realtime_augmentation.py:761

realtime_augmentation.py:784

Only the valid border mode is supported. n_filters should be a multiple of 16

cc_layers.py:29

pool_size is an INTEGER, not a tuple. We can only do square pooling windows. if the stride is none, it is taken to be the sa

cc_layers.py:105

This implements stochastic pooling as in Zeiler et al. 2013 to replace max pooling. Pooling is stochastic by default. When dropout_ac

cc_layers.py:140

(self, input_layer, alpha=1e-4, beta=0.75, size_f=5, blocked=True)

cc_layers.py:188

(self, input_layer)

cc_layers.py:217

(self, input_layer)

cc_layers.py:237

This is a convolution which is circular in the 0-direction, and valid in the 1-direction. n_filters should be a multiple of 16

cc_layers.py:255

pool_size: the number of inputs to be pooled together.

cc_layers.py:362

(self, mb_size, n_features, length)

(self, mb_size, n_features)

(self, mb_size, n_features, width, height)

(self, input_layer, ds_factor, ignore_border=False)

(self, input_layer, pool_size, ignore_border=False)

(self, input_layer, pooling_function='mean')

(self, input_layer, n_outputs, weights_std, init_bias_value, nonlinearity=rectify, dropout=0.)

(self, input_layer, n_filters, filter_length, weights_std, init_bias_value, nonlinearity=rectify, flip_conv_di

(self, input_layer, n_filters, filter_length, stride, weights_std, init_bias_value, nonlinearity=rectify, drop

(self, input_layer, n_filters, filter_width, filter_height, weights_std, init_bias_value, nonlinearity=rectify

(self, input_layer, n_filters_per_unit, dropout=0.)

(self, input_layer, n_outputs, weights_std, init_bias_value, nonlinearity=rectify, dropout=0., dropout_tied=Fa

(self, input_layer, n_filters_per_unit, dropout=0., pooling_function='max')

(self, input_layer, error_measure='mse')

(self, input_layer)

(self, input_layers)

n: window size k: bias alpha: scaling beta: power

implementation can be: - convolution: use conv2d with the subsample parameter - unstrided: use conv2d + reshaping so

(self, input_layer, part_size)

(self, input_layer)

(self, input_layers, part_size, include_flip=False)

num_views is the number of different input representations that were merged.

pool_size: the number of inputs to be pooled together. stride: the stride between pools, if not set it defaults to pool_size

pool_size: the number of inputs to be pooled together. feature_dim: the dimension of the input to pool across. By default this is 1

(self, input_layer)

(self, input_layer, threshold)

(self, input_layer)

(self, input_layer)

(self, input_layer)

(self, input_layer)

(self, input_layer)

(self, input_layer, scale=10.0)

(self, ds_transforms, augmentation_params, target_sizes=None)

realtime_augmentation.py:208

(self, subset, ds_transforms, augmentation_transforms, target_sizes=None)

realtime_augmentation.py:303

(self, ds_transforms, augmentation_params, target_sizes=None)

realtime_augmentation.py:558

(self, ds_transforms, augmentation_params, target_sizes=None)

realtime_augmentation.py:582

(self, ds_transforms, augmentation_params, target_sizes=None)

realtime_augmentation.py:606

(self, subset, ds_transforms, augmentation_transforms, target_sizes=None)

realtime_augmentation.py:649

(self, subset, ds_transforms, augmentation_transforms, target_sizes=None)

realtime_augmentation.py:677

(self, subset, ds_transforms, augmentation_transforms, target_sizes=None)

realtime_augmentation.py:705

(self, ds_transforms, augmentation_params, target_sizes=None)

realtime_augmentation.py:730

(self, subset, ds_transforms, augmentation_transforms, target_sizes=None)

realtime_augmentation.py:755

(self, func, *args, **kwargs)

realtime_augmentation.py:779

Function_buffered_generation_process

(source_gen, buffer)

load_data.py:553

Function_buffered_generation_thread

(source_gen, buffer)

load_data.py:517

Function_generation_thread

(source_gen, buffer, buffer_lock, buffer_size=2, sleep_time=1)

load_data.py:104

← previous next →101–200 of 336, ranked by callers