#Census API/Data Tools
##Purpose
This list is intended for people working with Census data in some kind of data management tool--SQL, Python, etc.
##Lists of Available Census APIs
Census Official
Secondary, more Detailed Census List in spreadsheet form
##Lists of Variables in each Census Product (e.g. ACS 2014)
http://api.census.gov/data/2014/acs5/variables.html
##Convert JSON (e.g. from API) to CSV convertcsv.com json-to-csv -- github page for project
citysdk
This package has a massively defined scope and seems to be under active development, but its not clear to me how and whether its ready for use.
Census2DBF
"Converts .csv files downloaded from the Census Factfinder into the legacy DBF format, which is useful for GIS applications. Adds a data dictionary to deal with field name issues(short strings)"
cenpy
An interface to explore and query the US Census API and return Pandas Dataframes. Ideally, this package is intended for exploratory data analysis and draws inspiration from sqlalchemy-like interfaces and acs.R.
example usage
addfips
Add a FIPS code to any CSV with a county name in it--seems extensible and useful.
census-us Basically, a list of FIPS codes for states--might be useful to have the same for MPOs, in several languages. Seems like in R or Python, it would be useful to just be able to import all the necessary FIPS codes for a query using county and/or city names.
census_area
This package has the potential to be useful but requires Shapely and therefore GDAL, which can be annoying to set up.
acs
The functionality of this package seems great--not sure about extensibility. Like the citysdk package, has a massive scope.
noncensus
tigris
US-Census 2000
census-cli
"Download 2014 American Community Survey 5-year estimates with population and income linked to a geojson at the tract level for a given zip code"
census-postgres
This package contains A LOT of work and understanding of the US Census information architecture. Here are a few choice quotes from the readme files in it:
####Flat File Organization:
"Each data product (e.g. American Community Survey 2006-2010) can be thought of as one large file, but the data are horizontally partitioned by state and are vertically separated into "segments" (in the Decennial Census) or "sequences" (in ACS) of less than 256 columns each..."
####Geographic Variables:
"It is expected that analysts will usually be working with data within a geographic scale (e.g. running a regression on county-level data), and often working within a defined region. Scale is identified by two columns, sumlevel and component, where sumlevel represents the level in the geographic hierarchy from Nation to State to County, etc. (and many others such as Tribal Tracts, ZIP Code Tabulation Areas or ZCTAs, Metropolitan and Micropolitan Areas, etc.) and component represents a subset of the population in that geography, e.g. Urban only or Rural only population. (The code 00 represents total population, and is often the component of interest. Codes other than 00 are only defined for county and geographies and larger.)"
Rather than require the analyst to constantly filter based on scale, the data may be partitioned by summary level and component.