You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A refactor in TopoStats (see issue) switched from using the topostats_object["image"] as input to grainstats to a newly introduced Class (ImageGrainCrops) which has two
attributes above and below which each hold a new class DirectionGrainCrops. The DirectionGrainCrops class has
two attributes, crops (a dictionary of GrainCrop classes which stores various properties about the grain such as the
mastk, image, bounding box padding and so forth) and the full_mask_tensor. The full mask tensors alone are stored in a topostats_object["grain_tensors"]["above"] and topostats_object["grain_tensors"]["below"] but at the moment the GrainCrop dictionaries are not saved, we will have to ensure they are saved to the HDF5 .topostats files on saving.
Because AFMReader should return the exact same object when loading a .topostats HDF5 file as that from which it
was saved we will have to ensure that all of this information is loaded and the appropriate classes are reconstructed so
that we can have entry points for grainstats, disordered_tracin, nodestats, ordered_tracing etc. to satisfy the
desire to have modular analyses available which shortens the processing and feedback loop when trying to improve
disentangling objects etc.
The text was updated successfully, but these errors were encountered:
A refactor in TopoStats (see issue) switched from using the
topostats_object["image"]
as input tograinstats
to a newly introduced Class (ImageGrainCrops
) which has twoattributes
above
andbelow
which each hold a new classDirectionGrainCrops
. TheDirectionGrainCrops
class hastwo attributes,
crops
(a dictionary ofGrainCrop
classes which stores various properties about the grain such as themastk, image, bounding box padding and so forth) and the
full_mask_tensor
. The full mask tensors alone are stored in atopostats_object["grain_tensors"]["above"]
andtopostats_object["grain_tensors"]["below"]
but at the moment theGrainCrop
dictionaries are not saved, we will have to ensure they are saved to the HDF5.topostats
files on saving.Because
AFMReader
should return the exact same object when loading a.topostats
HDF5 file as that from which itwas saved we will have to ensure that all of this information is loaded and the appropriate classes are reconstructed so
that we can have entry points for
grainstats
,disordered_tracin
,nodestats
,ordered_tracing
etc. to satisfy thedesire to have modular analyses available which shortens the processing and feedback loop when trying to improve
disentangling objects etc.
The text was updated successfully, but these errors were encountered: