Consuming Government Data with Python and D3
Pratap Vardhan (~pratap) |
58
Description:
The explosion of open data, especially government data has made it easy for us to see our own governance in a better light.
In this talk, We will analyze and visualize Indian Governance with the help of novel tools that include Python
, Pandas
and D3.js
At the end of this talk you will be able to,
- Identify, read, parse, ETL data
- Find patterns, anomalies in data
- Integrate d3.js to your python data pipeline
- Visualize patterns in a beautiful web/mobile friendly framework
Typical interactive dashboard will have features derived from these three.
- AAP Donations Tracker - An interactive dashboard to explore trends in donations received by the Aam Aadmi Party
- PM Speechopedia A visual gallery of the 1,198 speeches by Dr Manmohan Singh. Explorations of Textual co-occurrence networks
- Flags - Explore the flags of the world to see similarities in colors, symbols and patterns.
In the session, we would discuss, showcase the process and what I've learned from -- scraping data from sources, automating crawler jobs, storing structured data into DB, transforming data using Pandas, extracting features from dataset (Textual features from PM speeches, How Image features were extracted from flags).
And, for demonstration, the intent is to pick a dataset, preferably from https://data.gov.in/, which has scope for exploratory analysis. In the session, we would transform, create data views using Pandas
, spot patterns, and build an interactive dashboard in d3
, by integrating it into the Python
data pipeline.
Prerequisites:
- You are interested in either of Data, Governance, Analytics and Visualization.
- Python, Pandas, IPython installed on your system, if you want to play along.
Speaker Info:
Pratap Vardhan is a Data Scientist at Gramener.com, a Data Analytics and Visualization company.
Pratap has a B.Tech from NIT, Bhopal. He previously worked with Indian Statistical Institute, Indian Institute of Science and Thomson Reuters in the fields related to Computer Vision, HCI, Data Mining and Social Networks.
Speaker Links:
Talks:
- Data Visualization: Finding Pictures in Numbers at ISI-CODATA International Training Workshop on Big Data, 2015
- Exploratory Analysis of Indian Election Data at Open Data Camp-2014, ISB
- Analysis on Maharashtra State Elections-2014 written/covered in News Media
- Election math: What might have been, If Maharashtra’s alliances had remained in place - The Indian Express
- The Crest and Trough of a Wave - The Hindu
- Charts: Break-ups or not, BJP would have come out on top in Maharashtra - Firstpost
- Github repo for the analysis done on Maharashtra State Elections-2014 .