Popups on Choropleth Map

[mattwg]

Hi - I love what you have created here - today I manage to get an aggregate query from elastic search displayed on a Folium map in a few hours - it was really easy!

Now I am trying to add popups, but I cannot see any example that shows how to get a popup on a choropleth map - could someone point me in the right direction please.

I would just want to display the value associated with the polygon. I was thinking of adding the value back to the underlying geojson object and then adding that on as some form of transparent layer on top of the choropleth - but not having much luck!!

If you point me in the right direction I will create an example notebook for others as thanks ;-) My coding probably is not up to much more contribution at this stage!

Matt

[BibMartin]

Hi @mattwg ,

thanks for your feedback and engagement.
It seems @om-henners is working on this (see #376).

For short term concerns, you can get a first workaround (not perfect though) in http://nbviewer.jupyter.org/gist/BibMartin/4b9784461d2fa0d89353 But this solution is less serious as @om-henners's implementation.

[mattwg]

Hi Martin - thanks for the pointers I will take a look - and make sure I
follow #376 #376.
Matt

On Mon, Feb 29, 2016 at 3:04 AM, Martin Journois notifications@github.com
wrote:

Hi @mattwg https://github.com/mattwg ,

thanks for your feedback and engagement.
It seems @om-henners https://github.com/om-henners is working on this
(see #376 #376).

For short term concerns, you can get a first workaround (not perfect
though) in http://nbviewer.jupyter.org/gist/BibMartin/4b9784461d2fa0d89353
But this solution is less serious as @om-henners
https://github.com/om-henners's implementation.

—
Reply to this email directly or view it on GitHub
#375 (comment)
.

[Conengmo]

I think the GeoJsonTooltip class adds the functionality requested in this issue. It's available since 0.6.0. It doesn't work with the Map.choropleth method easily though, users have to manually get the GeoJson or TopoJson object created by Map.choropleth and add their GeoJsonTooltip to that.

Closing this one.

[jtbaker]

@Conengmo have you had a chance to look at #964?

[Conengmo]

Ah you're right, I'm mixing up tooltips and popups. Reopening. When we finish your (upcoming) PR, this can be closed.

I have looked at your issue, but not good enough yet for a reply. I'm flagging it so I'll remember to do that soon.

[araichev]

@Conengmo , regarding your comment

I think the GeoJsonTooltip class adds the functionality requested in this issue. It's available since 0.6.0. It doesn't work with the Map.choropleth method easily though, users have to manually get the GeoJson or TopoJson object created by Map.choropleth and add their GeoJsonTooltip to that.

could you elaborate on it with an example illustrating how to add a tooltip to a choropleth?
I'm trying to do that exact thing and failing.

[jtbaker]

@araichev, are you on the 0.6.0 release?

Check the docstring for GeoJsonTooltip to reference its functionality with the help command, it covers the functionality pretty well.

I have a a notebook example in my geojsonmarkers PR (not yet merged) that demonstrates it how the GeoJsonTooltip can be used. The marker argument is not yet active (I need to fix a few things) but the tooltipping will work the same way for a Choropleth/Polygon map.

You can see it here: https://nbviewer.jupyter.org/github/jtbaker/folium/blob/geojsonmarker/examples/GeoJsonMarkersandTooltips.ipynb

[araichev]

Thanks, @jtbaker .
I'm starting to understand, but am not all the way there yet.

Yes, i am on the 0.6.0 release.
I can create a choropleth with my tiles, GeoJSON layer, and colorscale.
I can create a tooltip on a copy of the original GeoJSON layer like you do in your notebook.
But how do i replace the choropleth GeoJSON layer with the second one that has the tooltip and still keep the choropleth colorscale?
Or am i going about this wrong?

[jtbaker]

This conditional styling is done using the style_function argument in GeoJson. It takes a function as an an argument, which will evaluate on each feature in the dataset, and return a value based on the logic you give it. For simple fill colors, its easiest to pass a lambda function and query the object properties directly.

Here's the example in the official documentation (check cells 9 through 13 or so): https://nbviewer.jupyter.org/github/python-visualization/folium/blob/master/examples/GeoJSON_and_choropleth.ipynb

I think it is just more straightforward to query directly against the value instead of creating a new series/dict, just remember you have to pass through the properties key to get to your actual key value.

branca, a sibling of folium that you should have as well makes it easier to cook up linear and step colorscales with the use of d3.

Some examples here:
https://nbviewer.jupyter.org/github/python-visualization/folium/blob/master/examples/Colormaps.ipynb

Check out the /examples directory on the main master branch for more usages and examples on how to work with the api. You can view them with output rendered on NBviewer: https://nbviewer.jupyter.org/github/python-visualization/folium/tree/master/examples/

Hope this helps, let me know if you have any other questions!