DM-44169: Fix compensated gaussian measurement error raising code #273
Conversation
|
It seems you should refactor your commits...I don't think leaving the first one is a good idea (i.e. you now know |
|
Refactored the commits! |
| # failure flags | ||
| failure_flag = flagDefs.add(f"{width}_flag", "Compensated Gaussian measurement failed") | ||
| oob_flag = flagDefs.add(f"{width}_flag_bounds", "Compensated Gaussian out-of-bounds") | ||
|
|
There was a problem hiding this comment.
I think width evaluates to to a number (integer). Do you think it will be apparent "enough" that this number in the flag name is the kernel width? And will the casual user be expected to translate bounds to outOfBounds internally (probably yes to both, but my ~10pm brain didn't jump there right away)?
There was a problem hiding this comment.
That's the name of the measurement ... base_compensatedGaussian_5 or whatever. It's what we decided on ages ago, and it may not even matter since the compensated gaussians are not working well as a calibration flux measurement. 😞
As for the bounds, this is also the name from before, except done correctly. I don't think this is going to be used for anything anyway.
| y_slice = slice(y_floor - rad, y_floor + rad + 1, 1) | ||
| x_slice = slice(x_floor - rad, x_floor + rad + 1, 1) | ||
| y_mean = y - y_floor + rad | ||
| x_mean = x - x_floor + rad | ||
|
|
||
| sub_im = exposure.image.array[y_slice, x_slice] | ||
| sub_var = exposure.variance.array[y_slice, x_slice] |
There was a problem hiding this comment.
You could technically move setting sub_var to after the size checking (but readability may win out to leaving here...up to you).
No description provided.