Add sample data set and a documentation about it

[adorilson]

Add sample data set and a documentation about it

[gabubellon]

A suggestion to separate a sample data:

• Get the most actual affiliations data (warning, this data has 2.9G when unzipped);
• Get the most actual candidates data;
• Separate a sample of affiliations, filter by a city and a year (may be Brasilia and 2018);
• Join the separated with candidates data by vote_id (titulo de eleitor) field to get only candidates from same city and year;
• Save both data to sample files affiliations_sample.csv and candidates_sample.csv