Computer Vision for Medical Imaging

Documentation for package ‘rplatypus’ version 0.2.1

Help Pages

A B C D E I L M O P R S T U V Y

-- A --

as.list.platypus_spec	Turn a specification back into plain R data
augmentation_step	An augmentation step
available_callbacks	What this package can do, asked of the package
available_components	What this package can do, asked of the package
available_losses	What this package can do, asked of the package
available_metrics	What this package can do, asked of the package
available_models	What this package can do, asked of the package
available_optimizers	What this package can do, asked of the package
available_tasks	What this package can do, asked of the package
available_transforms	Which augmentations are available
available_weights	The published weights, and what they are

-- B --

binary_colormap	Colormaps and their labels
binary_labels	Colormaps and their labels

-- C --

callbacks	Callbacks
callback_cosine_annealing	Callbacks
callback_csv_logger	Callbacks
callback_early_stopping	Callbacks
callback_model_checkpoint	Callbacks
callback_reduce_lr_on_plateau	Callbacks
callback_swa	Average the weights over the last part of the run
callback_terminate_on_nan	Callbacks
classes_to_colours	Paint a mask with its colormap
colormaps	Colormaps and their labels
colours_to_classes	Read a colour mask back to class indices
ct_windows	The named CT windows

-- D --

decode_connected_components	How predicted channels become objects
decode_follow_flows	How predicted channels become objects
decode_nearest_seed	How predicted channels become objects
decode_watershed	How predicted channels become objects
decodings	How predicted channels become objects
detection_anchors	The anchors a detector used, and whether they were fitted
detection_crops	Cut every detection out of the image it was found in
detection_data	Where the images and their annotation files are
drawing_style	The decisions a drawing makes

-- E --

encoders	Pretrained encoders, and what they are worth
evaluate	Compare the trained models
evaluate.platypus_fit	Compare the trained models
evaluate_cases	Score every case separately
evaluate_cases.platypus_fit	Score every case separately
evaluate_classes	Average precision per class
evaluate_classes.platypus_fit	Average precision per class
evaluate_images	Score every image separately
evaluate_images.platypus_fit	Score every image separately
export_weights	Save a trained model's weights

-- I --

inspect_masks	What the colormap or labels match in your masks
instance_data	Data for instance segmentation: tell the objects apart, not just find them

-- L --

linknet	Segmentation models
losses	Loss functions
loss_boundary	A loss that knows how far a wrong voxel is from the truth
loss_cce	Loss functions
loss_cce_dice	Loss functions
loss_combo	Loss functions
loss_dice	Loss functions
loss_focal	Loss functions
loss_focal_tversky	Loss functions
loss_iou	Loss functions
loss_l1	Loss functions
loss_lovasz	Loss functions
loss_mse	Loss functions
loss_smooth_l1	Loss functions
loss_tversky	Loss functions

-- M --

mask_coverage	How much of each class a mask covers
mask_volume	How much of a volume a class occupies, in millilitres
metrics	Metrics
metric_cldice	Metrics
metric_dice	Metrics
metric_iou	Metrics
metric_tversky	Metrics
models	Segmentation models

-- O --

optimizers	Optimisers
optimizer_adadelta	Optimisers
optimizer_adagrad	Optimisers
optimizer_adam	Optimisers
optimizer_adamax	Optimisers
optimizer_adamw	Optimisers
optimizer_nadam	Optimisers
optimizer_rmsprop	Optimisers
optimizer_sgd	Optimisers
overlay_agreement	Where a prediction and the truth disagree
overlay_mask	Lay a mask over the image it belongs to

-- P --

platypus_conditions	When something is wrong: the conditions this package raises
platypus_device	Where the work will happen
platypus_fit	Train the models in a specification
platypus_spec	Build an experiment specification
platypus_status	Is the engine ready?
platypus_use_torch	Ask for a particular PyTorch build
plot.platypus_fit	Plot what happened during training
plot_anchors	Draw the anchors against the boxes they were fitted to
plot_boxes	Draw boxes on images
plot_masks	Look at masks beside the images they came from
predict.platypus_fit	Predict masks

-- R --

read_images	Read images as the model sees them
read_masks	Read mask files as class indices
read_record	Read what a run wrote down about itself
res_u_net	Segmentation models

-- S --

save_masks	Write masks to files
save_spec	Write a specification back out as YAML
save_volumes	Save predicted volumes as NIfTI
segmentation_data	Where the images and masks are
separations	What the network predicts, so objects can be told apart
separation_flows	What the network predicts, so objects can be told apart
separation_rim	What the network predicts, so objects can be told apart
series_report	Check folders of DICOM slices before training on them
split_dataset	Split a dataset into training, validation and test sets
split_path	The files in one part of a split
summary.platypus_cases	Score every case separately

-- T --

training_history

The epoch-by-epoch record

-- U --

unite_masks	Combine several binary masks into one
u_net	Segmentation models
u_net_plus_plus	Segmentation models

-- V --

voc_colormap	Colormaps and their labels
voc_labels	Colormaps and their labels
volume_info	What is in a volume file

-- Y --

YOLOv3