ou-wbpy

Slight fork of / wrapper for wbpy for use in JupyterLite / OpenLearn Learn to Code for Data Analysis.

Usage:

from ou_wbpy.api import wbapi

api = wbapi()
api.get_dataframe("SP.POP.TOTL", date="2010:2012")

Uses pyodide load machinery in Pyodide JupyterLite kernel (detected by trying import pyodide, else uses normal Python url loaders).

Using pandas-datareader

I originally hacked ou-wbpy from reviewing all the World Bank API reading packages I found (none of which work using the JupyterLite Pyodide kernel out of the can) whilst looking for a replacement for the pandas-datareader package (which was originally pulled out of the main pandas package), which also doesn't work unpatched in Pyodide.

But - it does, with a patch...

#Install package and its requirements
%pip install pandas_datareader distutils

#Patch various common packages
try:
  import pyodide_http
  pyodide_http.patch_all()
except:
  pass

from pandas_datareader import wb

wb.download(indicator='NY.GDP.PCAP.KD', country=['US', 'CA', 'MX'], start=2005, end=2008)

Using wbpy natively

We can also path wbpy, noting as well we need to update the base URL:

%pip install wbpy
import pyodide_http
pyodide_http.patch_all()
import wbpy

api = wbpy.IndicatorAPI()

iso_country_codes = ["GB", "FR", "JP"]
total_population = "SP.POP.TOTL"
api.BASE_URL = "https://api.worldbank.org/v2/"
dataset = api.get_dataset(total_population, iso_country_codes, date="2010:2012")
dataset
```