add moran_facette functionality and merge esda.moran plots to moran_scatterplot
#27
Conversation
splot/_viz_value_by_alpha.py
Outdated
| __author__ = ("Stefanie Lumnitz <stefanie.lumitz@gmail.com>") | ||
|
|
||
|
|
||
| def value_by_alpha(x, alphay, gdf): |
There was a problem hiding this comment.
sorry, I should have added value_by_alpha functionality is dealt with separately in the next pr #28
splot/_viz_esda_mpl.py
Outdated
| @@ -1062,3 +1160,113 @@ def moran_loc_bv_scatterplot(moran_loc_bv, p=None, | |||
| ax.plot(lag, fit.predy, **fitline_kwds) | |||
| ax.scatter(moran_loc_bv.zy, fit.predy, **scatter_kwds) | |||
| return fig, ax | |||
|
|
|||
|
|
|||
| def moran_facette(moran_matrix, figsize=(16,12), | |||
splot/_viz_esda_mpl.py
Outdated
| sharey=True, sharex=True) | ||
| fig.suptitle('Moran Facette') | ||
|
|
||
| for row in range(nrows): |
There was a problem hiding this comment.
Just as a suggestion: would it make sense to either use seaborn.FacetGrid to make our facets more abstract, that way making them more re-usable across the library? You might have considered this already and it might make not sense but just checking here. Ideally, if we're gonna have facets in several places of splot, the more we can reuse, the easier it'll be to maintain.
There was a problem hiding this comment.
seaboard.FacetGrid takes as an input data : DataFrame Tidy (“long-form”) dataframe where each column is a variable and each row is an observation.
Since the Moran_BV_matrix returns Moran_BV objects I thought reusing the scatterplot, that plots Moran_BV objects is the way to go. Otherwise we would need a Moran_Matrix that returns spatially lagged values and values.
splot/_viz_esda_mpl.py
Outdated
|
|
||
| Returns | ||
| ------- | ||
| fig : Matplotlib Figure instance |
There was a problem hiding this comment.
Does it always return a fig object? Even when ax is not None? I couldn't tell, but usually, if you pass and ax I think usually you return only ax (not completely sure though).
There was a problem hiding this comment.
yes, it always returns a fig. If ax is not None, else: fig = ax.get_figure() will create a fig.
splot/_viz_value_by_alpha.py
Outdated
|
|
||
|
|
||
|
|
||
| def value_by_alpha(x='attribute', y='alpha_attribute', c, data=geodataframe): |
There was a problem hiding this comment.
This method name is picked in L.13, no?
There was a problem hiding this comment.
sorry, I should have added value_by_alpha functionality is dealt with separately in the next pr #28
splot/_viz_value_by_alpha.py
Outdated
| fig, ax = geodataframe.plot(x, color=rgba) | ||
| return fig, ax | ||
|
|
||
| def make_RGBA_array_from_attribute(x): |
There was a problem hiding this comment.
sorry, I should have added value_by_alpha functionality is dealt with separately in the next pr #28
splot/_viz_value_by_alpha.py
Outdated
|
|
||
| def value_by_alpha(x='attribute', y='alpha_attribute', c, data=geodataframe): | ||
| rgb = matplotlib.colors.to_rgba_array(c) | ||
| alpha_channel = x/y.max() |
There was a problem hiding this comment.
This is great! could we generalize it a bit? I'd like to be able to have:
- diverging scaling (extremes are opaque, middle is transparent, something like
(x-np.median(y)).abs() / np.abs(y).max() - logarithmic scaling (including log divergent), which changes over the log of
y(ory + y.min()rather thanydirectly, ifyhas negative elements)
There was a problem hiding this comment.
I can try add more functionality to #28 :)
slumnitz
force-pushed
the
facette
branch
2 times, most recently
from
d80065a to
d7ec801
Compare
|
Now, this is redundant, yes? |
|
Should the PR be closed or is it still in play? |
|
@jGaboardi Thank you this can be closed. |
moran_facet()with documentation, tests and examplesmoran_scatterplot()that can use allesda.moranobjects as inputwith documentation, tests and examples
This pull requests builds ontop of bug fixes in #25
and depends on the integration of
varnamesin esda pysal/esda#23