Python module for UK polling analysis

My previous post described a web-service for providing up-to-date UK political polling information.  I’ve created a Python client to make use of this to perform simple analysis.

This allows you quickly do analysis such as polling a monthly average of the parties polling results.

parties

This can be generated using code like this.

from datetime import datetime
import pandas as pd
from pollingclient.client import PollingClient

pc = PollingClient()
dfm = pc.monthly_average()
dfm = dfm[dfm.index>datetime(2010,1,1)]
color_map = {'Con': 'blue', 'Lab': 'red', 'LD': 'orange', 'Green': 'green', 'UKIP': 'purple'}
colors = [color_map[x] for x in dfm.columns]
dfm.plot(style=colors);

Or another example, list the months in which Labour had the lowest average poll rating while in opposition.

Con Lab LD UKIP Green
Feb 2017 41.1 25.9 10.4 12.6 3.9
Dec 1981 25.0 26.2 46.8 0.0 0.0
Jun 1982 48.3 26.3 24.0 0.0 0.0
Apr 2017 45.3 26.3 10.6 8.3 3.3
Sep 1983 45.5 26.7 26.3 0.0 0.0

And the code for this.

dfm = pc.monthly_average()
l_in = dfm[list(dfm.index.map(lambda x: not pc.party_in_power('Lab', x)))]
dfx = l_in.sort_values("Lab").round(1).head(5)
dfx.index = dfx.index.strftime('%b %Y')
print dfx

Installation

The code can be found here on github.

You can install it from there like this:

pip install -e git+https://github.com/NickAltmann/UKPollingClient#egg=pollingclient

Examples of use

There are some notebook files on github here which illustrate the interface and provide further examples of what can be done.

Source credit

Please see my previous post about crediting the sources of this data.

Posted in Polling, Python | Leave a comment