A placeholder quickmap product with all 0s, just to start producing s…#3350
A placeholder quickmap product with all 0s, just to start producing s…#3350vineetbansal wants to merge 1 commit into
Conversation
tmplummer
left a comment
tmplummer
left a comment
There was a problem hiding this comment.
I don't have much to comment on for detailed implementation. My comments should express my opinion that the quickmaps use a different descriptor and different overall L2 function.
| projected onto it, see ``_inputs_at_map_pivot_angle``. When none of the | ||
| psets match (or no psets were passed-in to begin with), a quickmap is | ||
| produced instead. | ||
|
|
There was a problem hiding this comment.
Is this trying to make this function do too much. Should there be a separate function for normal maps and quickmaps?
| class Lo(ProcessInstrument): | ||
| """Process IMAP-Lo.""" | ||
|
|
||
| def pre_processing(self) -> ProcessingInputCollection: |
There was a problem hiding this comment.
I think that the reason it was recommended to filter here was so that the Parents attribute would be correct when the CDF gets written. Is there a reason why filtering on angle here doesn't work?
| dataset = cleanup_intermediate_variables(dataset) | ||
|
|
||
| return [ | ||
| sky_map.build_cdf_dataset( |
There was a problem hiding this comment.
It seems like the descriptor should be modified somehow to indicate this is a quickmap? I would think that thsoe should be easily distinguished from regular maps by the filename.
| @@ -1323,18 +1275,21 @@ def do_processing( | |||
| datasets = lo_l1c.lo_l1c(data_dict, anc_dependencies) | |||
|
|
|||
| elif self.data_level == "l2": | |||
There was a problem hiding this comment.
Can you distinguish quickmaps by the descriptor? I would think that this is the place to branch and just call a different function in lo_l2.py1 that will make the quickmap.
|
I just ran this by Mitchell and learned something new - they've been informally calling these quickmaps all along because they under the 1yr cadence. So it's essentially "the" map and can use the official descriptor. For your other points - good point about the parents attribute - I had forgotten about that. I can move filtering to the pre_processing step. |
2 participants
Closes #3349
lo_l2()now receives all inputs at all pivot angles and calls the new_inputs_at_map_pivot_angle()helper, which selects inputs against the pivot angle encoded in the map descriptor using a tightenedPSET_PIVOT_ANGLE_TOLERANCEof 5 degrees. This tolerance matches what theloteam uses in their technical documentation (and is guaranteed not to cause overlaps for now).quickmap support: rather than raising
ValueErrorwhen no pointing sets are present,lo_l2()falls back to_lo_l2_quickmap(), which builds a correctly shaped, zero-filled rectangular ENA map from the descriptor alone. No psets are needed for the generation of quickmaps.The actual implementation is coming in another PR soon. In the meantime, the logger messages in here will allow us to see if the correct inputs are being received and filtered.