MCPcopy Create free account

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

Functions336 in github.com/benanne/kaggle-galaxies

⨍ Functions 336◇ Types & classes 53
↓ 51 callersMethoderror
(self, *args, **kwargs)
layers.py:836
↓ 50 callersMethodoutput
(self, input=None, *args, **kwargs)
layers.py:462
↓ 34 callersMethodpredictions
(self, *args, **kwargs)
layers.py:876
↓ 32 callersMethodget_output_shape
(self)
layers.py:457
↓ 12 callersFunctionshared_single
Shortcut to create an undefined single precision Theano shared variable.
layers.py:236
↓ 11 callersMethodgrad
(self, args, g_outs)
consider_constant.py:8
↓ 11 callersFunctiontc_softmax
(x, t)
custom.py:24
↓ 9 callersMethodget_output_shape
(self)
cc_layers.py:74
↓ 7 callersMethodoutput
(self, input=None, dropout_active=True, *args, **kwargs)
cc_layers.py:80
↓ 4 callersFunctionbuild_augmentation_transform
(zoom=1.0, rotation=0, shear=0, translation=(0, 0))
realtime_augmentation.py:66
↓ 4 callersFunctionfast_warp
This wrapper function is about five times faster than skimage.transform.warp, for our use case.
realtime_augmentation.py:47
↓ 3 callersFunctionall_parameters
Recursive function to gather all parameters, starting from the output layer
layers.py:46
↓ 3 callersFunctionaugment_fixed_and_dscrop_with_prepro
(img, ds_transforms, augmentation_transforms, target_sizes=None, prepro_transform=tform_identity)
realtime_augmentation.py:620
↓ 3 callersFunctionbuild_ds_transform
This version is a bit more 'correct', it mimics the skimage.transform.resize function.
realtime_augmentation.py:79
↓ 3 callersFunctionperturb_and_dscrop_with_prepro
This version supports a preprocessing transform which is applied before anything else
realtime_augmentation.py:526
↓ 3 callersMethodquestion_weights
q is a list of matrices of length 11 (one for each question), like the output given by targets() and answer_probabilities()
custom.py:516
↓ 3 callersMethodrescaling_updates
(self, c)
layers.py:415
↓ 2 callersMethodanswer_probabilities
normalise the answer groups for each question.
custom.py:450
↓ 2 callersFunctionaugment_fixed_and_dscrop
(img, ds_transforms, augmentation_transforms, target_sizes=None)
realtime_augmentation.py:276
↓ 2 callersFunctionbuild_pysex_center_rescale_transform
(img_index, subset='train', target_radius=160.0)
realtime_augmentation.py:487
↓ 2 callersFunctionbuild_pysex_center_transform
(img_index, subset='train')
realtime_augmentation.py:478
↓ 2 callersFunctionbuild_pysexgen1_center_rescale_transform
(img_index, subset='train', target_radius=160.0)
realtime_augmentation.py:506
↓ 2 callersFunctioncycle
(l, shuffle=True)
load_data.py:67
↓ 2 callersFunctionnumpy_loader_gen
(paths_gen, shuffle=True)
load_data.py:444
↓ 2 callersFunctionperturb_and_dscrop
(img, ds_transforms, augmentation_params, target_sizes=None)
realtime_augmentation.py:155
↓ 2 callersFunctionrandom_perturbation_transform
(zoom_range, rotation_range, shear_range, translation_range, do_flip=False)
realtime_augmentation.py:127
↓ 2 callersMethodreset_params
(self)
layers.py:453
↓ 1 callersFunctionall_bias_parameters
Recursive function to gather all bias parameters, starting from the output layer
layers.py:57
↓ 1 callersFunctionall_layers
Recursive function to gather all layers below the given layer (including the given layer)
layers.py:35
↓ 1 callersMethodanswer_probabilities
apply softmax functions to the answer groups for each question.
custom.py:64
↓ 1 callersMethodanswer_probabilities
apply softmax functions to the answer groups for each question.
custom.py:164
↓ 1 callersMethodanswer_probabilities
normalise the answer groups for each question.
custom.py:250
↓ 1 callersMethodanswer_probabilities
normalise the answer groups for each question.
custom.py:349
↓ 1 callersMethodanswer_probabilities
normalise the answer groups for each question.
custom.py:632
↓ 1 callersMethodanswer_probabilities
normalise the answer groups for each question.
custom.py:743
↓ 1 callersMethodanswer_probabilities
normalise the answer groups for each question.
custom.py:858
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_normconstraint.py:121
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_shareddense.py:121
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense_dup3.py:121
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_pysex.py:123
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_8433n_maxout2048.py:121
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_8433n_maxout2048_extradense_pysex.py:123
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense_pysexgen1_dup.py:123
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense.py:121
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense_big256.py:121
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense_pysex.py:123
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_8433n_maxout2048_extradense.py:121
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense_pysexgen1_dup2.py:123
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_pysex.py:123
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_shareddense512.py:121
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_8433n_maxout2048_pysex.py:123
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048.py:121
↓ 1 callersFunctioncreate_test_gen
()
try_convnet_cc_multirotflip_3x69r45_bigger.py:121
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_normconstraint.py:115
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_shareddense.py:115
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense_dup3.py:115
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_pysex.py:116
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_8433n_maxout2048.py:115
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_8433n_maxout2048_extradense_pysex.py:116
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense_pysexgen1_dup.py:116
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense.py:115
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense_big256.py:115
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense_pysex.py:116
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_8433n_maxout2048_extradense.py:115
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_extradense_pysexgen1_dup2.py:116
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048_pysex.py:116
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_shareddense512.py:115
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_8433n_maxout2048_pysex.py:116
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_maxout2048.py:115
↓ 1 callersFunctioncreate_valid_gen
()
try_convnet_cc_multirotflip_3x69r45_bigger.py:115
↓ 1 callersMethoderror
(self, *args, **kwargs)
custom.py:511
↓ 1 callersFunctionestimate_params
(img)
extract_pysex_params_gen2.py:22
↓ 1 callersFunctionestimate_params
(img)
extract_pysex_params_extra.py:22
↓ 1 callersMethodgenerate_normalisation_mask
when the clipped input is multiplied by the normalisation mask, the normalisation denominators are generated. So then we can just div
custom.py:611
↓ 1 callersMethodgenerate_normalisation_mask
when the clipped input is multiplied by the normalisation mask, the normalisation denominators are generated. So then we can just div
custom.py:722
↓ 1 callersMethodgenerate_normalisation_mask
when the clipped input is multiplied by the normalisation mask, the normalisation denominators are generated. So then we can just div
custom.py:837
↓ 1 callersMethodget_output_shape
(self)
cc_layers.py:299
↓ 1 callersMethodget_output_shape
(self)
cc_layers.py:375
↓ 1 callersMethodget_output_shape
(self)
layers.py:889
↓ 1 callersMethodget_output_shape
(self)
layers.py:1006
↓ 1 callersMethodget_output_shape
(self)
layers.py:1209
↓ 1 callersMethodget_output_shape
(self)
layers.py:1289
↓ 1 callersMethodget_output_shape
(self)
layers.py:1416
↓ 1 callersFunctionget_param_values
(layer)
layers.py:89
↓ 1 callersFunctionim_lcn_bias
LCN with an std bias to avoid noise amplification
load_data.py:321
↓ 1 callersFunctionim_rgb_to_yuv
(img)
load_data.py:302
↓ 1 callersFunctionim_yuv_to_rgb
(img)
load_data.py:305
↓ 1 callersFunctionload_and_process_image
(img_index, ds_transforms, augmentation_params, target_sizes=None)
realtime_augmentation.py:185
↓ 1 callersFunctionload_and_process_image_brightness_norm
(img_index, ds_transforms, augmentation_params, target_sizes=None)
realtime_augmentation.py:721
↓ 1 callersFunctionload_and_process_image_fixed
(img_index, subset, ds_transforms, augmentation_transforms, target_sizes=None)
realtime_augmentation.py:288
↓ 1 callersFunctionload_and_process_image_fixed_brightness_norm
(img_index, subset, ds_transforms, augmentation_transforms, target_sizes=None)
realtime_augmentation.py:739
↓ 1 callersFunctionload_and_process_image_fixed_pysex_centering
(img_index, subset, ds_transforms, augmentation_transforms, target_sizes=None)
realtime_augmentation.py:632
↓ 1 callersFunctionload_and_process_image_fixed_pysex_centering_rescaling
(img_index, subset, ds_transforms, augmentation_transforms, target_sizes=None)
realtime_augmentation.py:660
↓ 1 callersFunctionload_and_process_image_fixed_pysexgen1_centering_rescaling
(img_index, subset, ds_transforms, augmentation_transforms, target_sizes=None)
realtime_augmentation.py:688
↓ 1 callersFunctionload_and_process_image_pysex_centering
(img_index, ds_transforms, augmentation_params, target_sizes=None)
realtime_augmentation.py:543
↓ 1 callersFunctionload_and_process_image_pysex_centering_rescaling
(img_index, ds_transforms, augmentation_params, target_sizes=None)
realtime_augmentation.py:567
↓ 1 callersFunctionload_and_process_image_pysexgen1_centering_rescaling
(img_index, ds_transforms, augmentation_params, target_sizes=None)
realtime_augmentation.py:591
↓ 1 callersFunctionload_gz
(path)
load_data.py:427
↓ 1 callersFunctionload_image
(img_id, subset='train', normalise=True, from_ram=False)
load_data.py:55
↓ 1 callersMethodnormreg
direct_weighting: if True, the weighting is applied directly to the constraints (before the barrier function). if False, all constrai
custom.py:537
next →1–100 of 336, ranked by callers