Data Visualization with Python Folium Maps

Well… using geospatial analysis, I could create a map where cities with higher traffic AND higher bike scores were shown.

Even better, changing attributes such as the color and size of map markers means that I can use one figure to visualize all three dimensions.

To create this map, I first installed the Python Folium library.

!conda install -c conda-forge folium=0.


0 –yes import folium import webbrowserThe first dimension I wanted to show is geolocation.

I set my map’s latitude and longitude based on the results of a simple Google search for “latitude and longitude of the US”.

Then, I defined my map (cleverly named traffic_map) to initialize at these coordinates.

The default zoom is set to five (I had to play around with this parameter until I found a good display).

latitude = 37.

0902longitude = -95.

7129traffic_map = folium.

Map(location=[latitude, longitude], zoom_start=5)The second dimension wanted to show is traffic congestion.

I grouped my traffic variable (“traffic_index”) into quartiles:df['traffic_index_quartile'] = pd.

qcut(df['traffic_index'], 4, labels=False)Next, I created a dictionary for the colors of my markers.

I decided on a range of primary colors from lightblue (0) to red (3).

I wanted the highest quartile (i.


the most congested cities) to stand out with a bold red, and the lowest quartile (i.


the least congested and least viable candidates) to more or less fade into the base-map.

colordict = {0: 'lightblue', 1: 'lightgreen', 2: 'orange', 3: 'red'}Side note: I found viable color names from this Stack Overflow post:https://stackoverflow.


I have tried most of the colors listed, and they have accurately displayed as their names suggest, except for “light red”, which either causes an error or shows up as black.

The third and final dimension on my map is city “bikeability”.

To visualize this dimension, I set the size of the city markers to be 0.

15 time the city’s bikescore (another parameter I had to play around with using trial and error until it “looked” right).

The complete code for the final map is as follows:for lat, lon, traffic_q, traffic, bike, city in zip(df['latitude'], df['longitude'], df['traffic_index_quartile'], df['traffic_index'], df['bike_score'], df['city']): folium.

CircleMarker( [lat, lon], radius=.

15*bike, popup = ('City: ' + str(city).

capitalize() + '<br>' 'Bike score: ' + str(bike) + '<br>' 'Traffic level: ' + str(traffic) +'%' ), color='b', key_on = traffic_q, threshold_scale=[0,1,2,3], fill_color=colordict[traffic_q], fill=True, fill_opacity=0.

7 ).

add_to(traffic_map)traffic_mapThrough this geospatial data visualization, we can quickly see two facts about cities and biking.

First, “bikeability” is positively related to traffic congestion.

Second, cities with high traffic AND “bikeability” (i.


large red markers) are more likely to be found in the Northeast, California, and Pacific Northwest regions of the U.


(with a few notable exceptions).

Check out my post about the full project here.

And feel free to leave feedback or thoughts in the comments!.. More details

Leave a Reply