Rcal 406 science array quantities #616
Conversation
…o ArrayQuanitites
github-actions
bot
added
Dark Current
dependencies
romancal/flatfield/flat_field.py
Outdated
| ru.electron ** 2 / u.s ** 2, dtype=science.data.value.dtype) | ||
|
|
||
| err_sqrt = np.sqrt(science.var_poisson.value + science.var_rnoise.value + science.var_flat) | ||
| science.err = u.Quantity(err_sqrt, ru.electron / u.s, dtype=err_sqrt.dtype) |
There was a problem hiding this comment.
I feel like we don't need to touch most of this section, since flat_data is unitless so things like science.data /= flat_data should do the right thing? What am I missing? One would only need to touch the except: AttributeError stanza, which was deleted here?
| 'weights': np.squeeze(weights), | ||
| 'crmag': crmag | ||
| 'crmag': u.Quantity(crmag, ru.electron, dtype=pedestal.dtype), | ||
| } |
There was a problem hiding this comment.
If we added these units back at the opt_info stage instead of here, we wouldn't need to repeat adding these units in create_image_model?
| out_model.groupdq = out_data.groupdq[0] | ||
| # Roman 2D | ||
| out_model.pixeldq = out_data.pixeldq | ||
| out_model.err = out_data.err[0] | ||
| out_model.err = u.Quantity(out_data.err[0], ru.DN, dtype=out_data.err.dtype) |
There was a problem hiding this comment.
Staring at this; does it make more sense to add these units onto out_data in process(...) following the do_correction call? Harping on my usual imagination that our ideal model should be: units everywhere, except around the calls to stcal, where there's a step beforehand removing units and afterward adding them back?
Resolves RCAL-406
This PR changes science products ndarrays to quantities.
Checklist
CHANGES.rstunder the corresponding subsection