💾 Load data
I've built a bot that harvested every tweet containing the hashtags #surf, #kitesurf and #windsurf for a couple of days. Those tweets all have a location with geographic coordinates. You can read more about this side project here.
The dataset is stored on github. Let's load it using pandas. Note that I've already aggregated the dataset per location. So for each location, I have a n column that tells the number of tweets.
# Libraries
import pandas as pd
# read the data (on the web)
data = pd.read_csv('https://raw.githubusercontent.com/holtzy/The-Python-Graph-Gallery/master/static/data/TweetSurfData.csv', sep=";")
# Check the first 2 rows
data.head(2)
| homelon | homelat | homecontinent | n | |
|---|---|---|---|---|
| 0 | -178.12 | -14.29 | Australia | 10 |
| 1 | -172.10 | -13.76 | NaN | 2 |
📍 Background map
As explained in the background map section of the gallery, there are several way to build a background map with Python. Here I suggest to use the basemap library that provides boundaries for every country:
# Basemap library
from mpl_toolkits.basemap import Basemap
import matplotlib.pyplot as plt
# Set the dimension of the figure
plt.rcParams["figure.figsize"]=15,10;
# Make the background map
m=Basemap(llcrnrlon=-180, llcrnrlat=-65, urcrnrlon=180, urcrnrlat=80, projection='merc');
m.drawmapboundary(fill_color='#A6CAE0', linewidth=0);
m.fillcontinents(color='grey', alpha=0.3);
m.drawcoastlines(linewidth=0.1, color="white");
⭕ Bubble map
Let's add each data point on the map thanks to the scatter() function. The scatter() funcion is described extensively in the scatterplot section of the gallery. x and y coordinates are longitude and latitude respectively. s is the size of each circle, it is mapped to the n color of the data frame.
# Make the background map
m=Basemap(llcrnrlon=-180, llcrnrlat=-65, urcrnrlon=180, urcrnrlat=80)
m.drawmapboundary(fill_color='#A6CAE0', linewidth=0)
m.fillcontinents(color='grey', alpha=0.3)
m.drawcoastlines(linewidth=0.1, color="white")
# prepare a color for each point depending on the continent.
data['labels_enc'] = pd.factorize(data['homecontinent'])[0]
# Add a point per position
m.scatter(
x=data['homelon'],
y=data['homelat'],
s=data['n']/6,
alpha=0.4,
c=data['labels_enc'],
cmap="Set1"
)
# copyright and source data info
plt.text( -175, -62,'Where people talk about #Surf\n\nData collected on twitter by @R_Graph_Gallery during 300 days\nPlot realized with Python and the Basemap library', ha='left', va='bottom', size=9, color='#555555' );
Note: I can't use the
mercatorprojection here. When I do, the circles coordinates are not recognized properly anymore. Please let me know if you have a fix!
