Declarative/simplified plotting #989
Conversation
dopplershift
added
Type: Feature
dopplershift
force-pushed
the
declarative
branch
4 times, most recently
from
6e07b54 to
da7ba5d
Compare
jrleeman
left a comment
jrleeman
left a comment
There was a problem hiding this comment.
Few minor comments/suggestions
examples/plots/Combined_plotting.py
Outdated
| # Plot the data on a map | ||
| panel = MapPanel() | ||
| panel.area = 'us' | ||
| panel.maps = ['coastline', 'borders', 'states', 'rivers', 'ocean', 'land'] |
There was a problem hiding this comment.
Maybe map layers? maps seems odd as we're making a map, and these are features/layers/things on the map.
| @@ -0,0 +1,31 @@ | |||
| # Copyright (c) 2019 MetPy Developers. | |||
There was a problem hiding this comment.
Maybe use a less generic name for this example? Simple image plot? Something that doesn't take a name with little description of the output.
metpy/plots/declarative.py
Outdated
|
|
||
| _projections = {'lcc': ccrs.LambertConformal(central_latitude=40, central_longitude=-100, | ||
| standard_parallels=[30, 60]), | ||
| 'ps': ccrs.NorthPolarStereo(central_longitude=-100), 'mer': ccrs.Mercator()} |
There was a problem hiding this comment.
Wrap each one to a new line? Just a style comment.
metpy/plots/declarative.py
Outdated
| return self.projection | ||
|
|
||
| @property | ||
| def _map_features(self): |
There was a problem hiding this comment.
Again - layers here maybe?
There was a problem hiding this comment.
internal, so less important, but with the rename of the public property, I'll make _layer_features.
|
|
||
|
|
||
| @exporter.export | ||
| class ImagePlot(Plot2D): |
There was a problem hiding this comment.
What if we have pure x/y data that we want to use with this syntax, not on a map? (similar to just wanting to make a line plot or similar. Maybe this is a image "heat map" of time series data.)
There was a problem hiding this comment.
I'm not convinced this will ever be something we really want to support, but we can revisit in the future if a compelling use case appears. I don't think we're doing anything that would make it difficult to support (heck, it might actually work currently).
There was a problem hiding this comment.
I may be biased, but one good use case could be cross sections! Or, would that work better as a separate class?
There was a problem hiding this comment.
Darn you and your good ideas @jthielen 😉
Will need to wait for next release, but that's a really good one. We'll need a new panel type for that as well.
| contours = Union([List(Float()), Int()], default_value=25) | ||
|
|
||
| @observe('contours', 'linecolor', 'linewidth') | ||
| def _set_need_rebuild(self, _): |
There was a problem hiding this comment.
Is this different from need redraw? Don't see that getting set here anyway.
There was a problem hiding this comment.
So the name of the method doesn't really matter as far as overriding, whatnot. I've been using "redraw" conceptually as just needing to call draw, which will redisplay with the current set of attributes. A "rebuild" means you need to re-execute the matplotlib plotting methods to replace the existing plot.
dopplershift
force-pushed
the
declarative
branch
2 times, most recently
from
56e0754 to
5aef138
Compare
Adds a declarative interface for creating plots, similar in nature to GEMPAK.
dopplershift
force-pushed
the
declarative
branch
6 times, most recently
from
a96e851 to
c370683
Compare
dopplershift
force-pushed
the
declarative
branch
from
c370683 to
e16c0b5
Compare
3 participants
First cut at the so-called "GEMPAK" style plotting interface. This provides a declarative interface for plotting, where the focus is on setting attributes to certain values. This also is wired to hopefully work properly with widgets in the notebook to provide a nice path to interactive, event-driven simple interfaces.
Still todo for this cut: