janbrus / stats-to-pandas

Download data from statistical agencies to pandas dataframes.

stats-to-pandas

Download data from statistical agencies to pandas dataframes.

Demo on YouTube

Installation

pip install stats_to_pandas

Requirements

• pandas
• requests
• pyjstat
• Jupyter notebook, IPython, ipywidgets

Overview

0. Import package

import stats_to_pandas as stp

1. Search for tables based on keywords

stp.search('cows')

2. Select a table id from the dataframe and create a gui-box to select the variables to be included

(Works when the user is in a jupyter notebook and has widgets installed)

box = stp.select(table_id = '10714')

3. Read the variable selection from the gui-box and download the data from Statistics Norway

df = stp.read_box(box)

4. More

Read all values for all variables for table with id 10714

(Avoid having to select variables, step 2 above, but may result in large tables)

df = stp.read_all(table_id = '10714')

Get the json string for a full query of table number 10714

query = stp.full_json(table_id = '10714', out = 'str')

Get the json dict for a full query of table number 10714

query = stp.full_json(table_id = '10714', out = 'dict')

Get the json dict associated with the current selection from a widget box

query = stp.get_json(box)

Read data from Statistics Ireland

irl_url = 'http://www.cso.ie/StatbankServices/StatbankServices.svc/jsonservice/responseinstance/CNA31'

df = stp.read_url(full_url = irl_url)