Clean up example scripts #26
Conversation
| xt[idx] = np.random.rand(s) | ||
| y = D @ xt + 5e-2 * np.random.randn(m) # synthetic signal | ||
| y = D @ xt + 5e-2 * np.random.randn(m) # Synthetic signal |
There was a problem hiding this comment.
Suggestion: update style guide to note this convention. Also update style guide to follow this convention.
| @@ -76,14 +77,16 @@ def demosaic(cfaimg): | |||
| """ | |||
| s = Afn(img) | |||
| rgbshp = s.shape + (3,) # Shape of reconstructed RGB image | |||
| nsigma = 2e-2 # Noise standard deviation | |||
| σ = 2e-2 # Noise standard deviation | |||
There was a problem hiding this comment.
Is this a style guideline? I hope not: I would have to search the web to discover how to type these characters in my editor, I imagine the same is true for others. I don't think the "cool factor" is worth it.
There was a problem hiding this comment.
No, it's not a style guideline, but it's already used quite extensively in the examples. Agreed on the extra pain in entering the characters, but I think it does enhance readability.
There was a problem hiding this comment.
Emacs-- C-x 8 Enter should open up the unicode search list, or M-x insert-char.
Our style guide says no unicode in public APIs. They are allowed in internal APIs. The user should never have to use unicode chars.
In this case, the user can choose to use unicode symbols or not; they aren't required.
|
|
||
| In SCICO, switching between these two regularizers is a one-line change: | ||
| replacing an [L1Norm](../_autosummary/scico.functional.rst#scico.functional.L1Norm) with a [L21Norm](../_autosummary/scico.functional.rst#scico.functional.L21Norm). | ||
| where $R$ is the isotropic TV: the sum of the norms of the gradient vectors at each point in the image $\mathbf{x}$. The same reconstruction is performed with anisotropic TV regularization for comparison; the isotropic version shows fewer block-like artifacts. |
There was a problem hiding this comment.
This is another issue that should be addressed in the style guide. At the moment, docstrings in examples often have very long lines. One could argue that it doesn't matter because they are primarily going to be viewed in markup-rendered form.
| @@ -34,7 +29,7 @@ | |||
| """ | |||
| N = 256 # Image size | |||
|
|
|||
| # these steps create a ground truth image by spatially filtering noise | |||
| # These steps create a ground truth image by spatially filtering noise | |||
There was a problem hiding this comment.
Example comments tend to be imperative, this isn't. Suggestion:
"Create a ground truth image by spatially filtering noise." (note full stop)
There was a problem hiding this comment.
Agreed. There are probably other examples like this that I also didn't catch,
| """ | ||
| Denoise with isotropic total variation | ||
| """ | ||
| reg_weight_iso = 2e0 | ||
| f = loss.SquaredL2Loss(y=y) | ||
| g_iso = reg_weight_iso * functional.L21Norm() | ||
|
|
||
| # the append=0 option makes the results of horizontal and vertical finite differences | ||
| # The append=0 option makes the results of horizontal and vertical finite differences |
There was a problem hiding this comment.
"...differences." (full stop)
| """ | ||
| Denoise with anisotropic total variation for comparison. | ||
| """ | ||
| # we tune the weight to give the same data fidelty as the isotropic case | ||
| # We tune the weight to give the same data fidelty as the isotropic case |
There was a problem hiding this comment.
"Tune the weight..." full stop
| fig.show() | ||
|
|
||
| # zoomed version | ||
| # Zoomed version |
examples/scripts/pgm_tv_isotropic.py
Outdated
| @@ -38,7 +34,8 @@ | |||
| """ | |||
| N = 256 # Image size | |||
|
|
|||
| # these steps create a ground truth image by spatially filtering noise | |||
|
|
|||
| # These steps create a ground truth image by spatially filtering noise | |||
There was a problem hiding this comment.
"Create a ground truth..." full stop.
Clean up example scripts