Getting KeyError after using "bins" parameter in Choropleth

[emontarsolo]

I tried to do a Choropleth map using Folium. The result was ok but I wanted the colormap to be more "shaded", since by default there is only 6 bins.
However when I set the parameter "bins=10", I get the error "KeyError: 'YlGn' "
I noticed I have the same error, when I specify the "bins" parameters in the "Choropleth maps" example of the QuickStart. So I will use this example to illustrate my issue.

The following code:

import folium
import pandas as pd


url = 'https://raw.githubusercontent.com/python-visualization/folium/master/examples/data'
state_geo = f'{url}/us-states.json'
state_unemployment = f'{url}/US_Unemployment_Oct2012.csv'
state_data = pd.read_csv(state_unemployment)

m = folium.Map(location=[48, -102], zoom_start=3)

folium.Choropleth(
    geo_data=state_geo,
    name='choropleth',
    data=state_data,
    columns=['State', 'Unemployment'],
    key_on='feature.id',
    fill_color='YlGn',
    fill_opacity=0.7,
    line_opacity=0.2,
    legend_name='Unemployment Rate (%)',
    bins=10
).add_to(m)

folium.LayerControl().add_to(m)

m

Leads to this error :

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
<ipython-input-6-d2bc56d78020> in <module>
     20     line_opacity=0.2,
     21     legend_name='Unemployment Rate (%)',
---> 22     bins=10
     23 ).add_to(m)
     24 

/usr/local/lib/python3.6/site-packages/folium/features.py in __init__(self, geo_data, data, columns, key_on, bins, fill_color, nan_fill_color, fill_opacity, nan_fill_opacity, line_color, line_weight, line_opacity, name, legend_name, overlay, control, show, topojson, smooth_factor, highlight, **kwargs)
   1088             # We add the colorscale
   1089             nb_bins = len(bin_edges) - 1
-> 1090             color_range = color_brewer(fill_color, n=nb_bins)
   1091             self.color_scale = StepColormap(
   1092                 color_range,

/usr/local/lib/python3.6/site-packages/branca/utilities.py in color_brewer(color_code, n)
    158     if not explicit_scheme:
    159         # Check to make sure that it is not a qualitative scheme.
--> 160         if scheme_info[base_code] == 'Qualitative':
    161             matching_quals = []
    162             for key in schemes:

KeyError: 'YlGn'

The same code without the " bins=10" works perfectly.

What Am I doing wrong here ? Is it because the color_brewer schemes only take 6 values ? What Am I supposed to do then ?

Python version : 3.6.8
Folium Version : '0.10.1'

[Conengmo]

Weird. bins accepts values higher than 6, that shouldn't be a problem. The error is weird, 'YlGn' isn't a key in scheme_info, though if you look in the code that key is there. The part that does this lookup only is triggered when there are more than 6 bins, so that explains why you didn't get the error with 6 bins.

[emontarsolo]

@Conengmo Thank you for your answer.

Sorry I forgot to give an update ! I found a solution to get around this issue: I use the GeoJson feature instead of the Choropeth feature. Since GeoJson allows you to use custom colormap, and even a continuous colormap, you don't have to set a number of bins. This example testing different colormaps and also this example using colormap + tooltip help me a lot.

I'm not sure I should close the issue since the bug is still here ?

[Conengmo]

Hi @emontarsolo, thanks for reporting back, happy to hear you got it working using GeoJson. Let's leave the issue open, me or someone else should check if this bug indeed exists and in that case what should be fixed.

[iacobo]

Weird. bins accepts values higher than 6, that shouldn't be a problem. The error is weird, 'YlGn' isn't a key in scheme_info, though if you look in the code that key is there. The part that does this lookup only is triggered when there are more than 6 bins, so that explains why you didn't get the error with 6 bins.

@Conengmo the problem is in branca/utilities.py, not folium/utilities.py:

/usr/local/lib/python3.6/site-packages/branca/utilities.py in color_brewer(color_code, n)

within the definiton of color_brewer:

def color_brewer(color_code, n=6):
    #...
    with open(os.path.join(rootpath, '_cnames.json')) as f:
        scheme_info = json.loads(f.read())

    #...

    try:
        schemes[core_color_code]
        explicit_scheme = True
    except KeyError:
        explicit_scheme = False

    # Only if n is greater than the scheme length do we interpolate values.
    if not explicit_scheme:
        # Check to make sure that it is not a qualitative scheme.
        if scheme_info[base_code] == 'Qualitative':
            matching_quals = []
            for key in schemes:
                if base_code + '_' in key:
                    matching_quals.append(int(key.split('_')[1]))

            raise ValueError('Expanded color support is not available'
                             ' for Qualitative schemes; restrict the'
                             ' number of colors for the ' + base_code +
                             ' code to between ' + str(min(matching_quals)) +
                             ' and ' + str(max(matching_quals))
                             )

The problem is:

1. none of the keys of '_cnames.json' have a value 'Qualitative'
2. the keys of '_cnames.json' are all individual colour names (e.g. 'indigo'), not colour-scheme names (e.g. 'YlGn').

[Conengmo]

Hi @ja-ox, thanks for looking into this! You're right, this is currently broken in branca.

Here we use the palette name and the number of bins to create a new name, which may not exist: https://github.com/python-visualization/branca/blob/master/branca/utilities.py#L128 We should make sure the integer we attach to the name results in an existing name.

A second problem is on this line: https://github.com/python-visualization/branca/blob/master/branca/utilities.py#L152 We should make this lookup robust, i.e. it doesn't matter if it's not in that dictionary, because then it's not Qualitative.

Is this something you want to help with? Or were you just pointing it out?

[ishant001]

I tried to do a Choropleth map using Folium. The result was ok but I wanted the colormap to be more "shaded", since by default there is only 6 bins.
However when I set the parameter "bins=10", I get the error "KeyError: 'YlGn' "
I noticed I have the same error, when I specify the "bins" parameters in the "Choropleth maps" example of the QuickStart. So I will use this example to illustrate my issue.

The following code:

import folium
import pandas as pd


url = 'https://raw.githubusercontent.com/python-visualization/folium/master/examples/data'
state_geo = f'{url}/us-states.json'
state_unemployment = f'{url}/US_Unemployment_Oct2012.csv'
state_data = pd.read_csv(state_unemployment)

m = folium.Map(location=[48, -102], zoom_start=3)

folium.Choropleth(
    geo_data=state_geo,
    name='choropleth',
    data=state_data,
    columns=['State', 'Unemployment'],
    key_on='feature.id',
    fill_color='YlGn',
    fill_opacity=0.7,
    line_opacity=0.2,
    legend_name='Unemployment Rate (%)',
    bins=10
).add_to(m)

folium.LayerControl().add_to(m)

m

Leads to this error :

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
<ipython-input-6-d2bc56d78020> in <module>
     20     line_opacity=0.2,
     21     legend_name='Unemployment Rate (%)',
---> 22     bins=10
     23 ).add_to(m)
     24 

/usr/local/lib/python3.6/site-packages/folium/features.py in __init__(self, geo_data, data, columns, key_on, bins, fill_color, nan_fill_color, fill_opacity, nan_fill_opacity, line_color, line_weight, line_opacity, name, legend_name, overlay, control, show, topojson, smooth_factor, highlight, **kwargs)
   1088             # We add the colorscale
   1089             nb_bins = len(bin_edges) - 1
-> 1090             color_range = color_brewer(fill_color, n=nb_bins)
   1091             self.color_scale = StepColormap(
   1092                 color_range,

/usr/local/lib/python3.6/site-packages/branca/utilities.py in color_brewer(color_code, n)
    158     if not explicit_scheme:
    159         # Check to make sure that it is not a qualitative scheme.
--> 160         if scheme_info[base_code] == 'Qualitative':
    161             matching_quals = []
    162             for key in schemes:

KeyError: 'YlGn'

The same code without the " bins=10" works perfectly.

What Am I doing wrong here ? Is it because the color_brewer schemes only take 6 values ? What Am I supposed to do then ?

Python version : 3.6.8
Folium Version : '0.10.1'

make bins=9
color YlGn has 9 shades

[lorenzoromani1983]

Hi. Is this fixed? I can only plot maps up to 9 bins. More than 9 bins still gives keyerror.

[Conengmo]

AFAIK this hasn't been fixed yet. We need a volunteer to work on this.

[Conengmo]

I believe this was fixed with python-visualization/branca#115, which was included in branca 0.6.0.