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
The development of an SJICubeSequence class would provide a few important benefits. Firstly, it would enable multiple SJI files to be read into a single object, say if a OBS was split over two files due an eclipse, or if a user wanted to combine data from different campaigns using the same OBS number. Secondly, it would enable an SJICube to be broken out into a sequence of single images, each with its own WCS. This would enable fine-scale pointing corrections to be made on a frame by frame basis, but still give users an object that they could use as a single cube.
Proposed Implementation
SJICubeSequence would be subclassed from NDCubeSequence
As SJI data is always 3D (unlike spectrograph data), the __getitem__ method would be over-ridden. If the dimensionality of the sequence was 3D, i.e. all sub-cubes were 2D, normal slicing would be implemented. If the dimensionality is 4D, i.e. sub-cubes are 3-D, the index_as_cube slicing property would be called. This could also be determined by whether there was a common axis set.
A new section of the read_iris_sji_level2_fits would be added to handle multiple files and return a SJICubeSequence.
The text was updated successfully, but these errors were encountered:
The development of an
SJICubeSequence
class would provide a few important benefits. Firstly, it would enable multiple SJI files to be read into a single object, say if a OBS was split over two files due an eclipse, or if a user wanted to combine data from different campaigns using the same OBS number. Secondly, it would enable anSJICube
to be broken out into a sequence of single images, each with its own WCS. This would enable fine-scale pointing corrections to be made on a frame by frame basis, but still give users an object that they could use as a single cube.Proposed Implementation
SJICubeSequence
would be subclassed fromNDCubeSequence
__getitem__
method would be over-ridden. If the dimensionality of the sequence was 3D, i.e. all sub-cubes were 2D, normal slicing would be implemented. If the dimensionality is 4D, i.e. sub-cubes are 3-D, theindex_as_cube
slicing property would be called. This could also be determined by whether there was a common axis set.read_iris_sji_level2_fits
would be added to handle multiple files and return aSJICubeSequence
.The text was updated successfully, but these errors were encountered: