Add PIXTLM and BGDTYP #90
Conversation
|
Also, doesn't look to be working at all on the ASVT data ignoring the pixel issue, so this is super WIP. |
|
I just added functionality to get VCDU counter. I also updated the test after adding VCDU counter. There are still no unit tests for the decom, but I looked at the test data from mark. I used this notebook: https://nbviewer.jupyter.org/urls/cxc.cfa.harvard.edu/mta/ASPECT/jgonzalez/Dynamic_Background_Decom.ipynb Note that the alternating raw and dynamic background values seems to start earlier than what Mark said. |
|
After the latest commit, I think this is ready. |
|
I think #89 is not controversial, since it is just passing through the kwargs. That is why I though of merging together with this. |
|
maude kwargs are used by the changes in this PR, because of the 'channel' argument. I would not add tests for that. |
|
OK. So right, the kwargs get "exercised" in that test anyway so this is done. |
|
I can't approve my own PR though and this is now half @javierggt 's work. @taldcroft do you want to approve before we merge? |
chandra_aca/maude_decom.py
Outdated
| @@ -355,9 +355,18 @@ def unpack_aca_telemetry(packet): | |||
| img_pixels = np.sum(np.packbits(_pixel_bits, axis=1) * [[2 ** 8, 1]], axis=1) | |||
| img_header['pixels'] = img_pixels | |||
| slots.append(img_header) | |||
| integ, glbstat = _unpack('>HB', packet[:3]) | |||
|
|
|||
| # The first two bytes that were integration time will have first two bits for | |||
There was a problem hiding this comment.
Write this comment so it will also work in 2027. You can mention pre- / post-dynamic background patch.
There was a problem hiding this comment.
I don't quite understand the comment on the comment about 2027, but seems fine to remove the "were integration time" part.
There was a problem hiding this comment.
I went with
# Before the dynamic background patch, the first two bytes contained INTEG in those
# 16 bits (named integbits). After the dynamic background patch, the first 6 bits of
# integbits will be repurposed: two bits for PIXTLM, next bit for BGDTYP, 3 spares,
# and 10 bits for INTEG. This telem/decom change is back-compatible and can be promoted
# before the dynamic background patch is in use onboard.
|
It looks like this was sufficiently reviewed and approved, so merging now. |
This includes the follwing changes: