"The U.S. Census Bureau's American Community Survey (ACS) is an ongoing survey that provides vital information on a yearly basis about our nation and its people. Information from the survey generates data that help determine how more than $400 billion in federal and state funds are distributed each year."
[https://www.census.gov/programs-surveys/acs/about.html]
Data from the ACS are published annually, aggregated as 1-year, 3-year, and 5-year estimates. Data from the survey are released at various geographical levels, the most detailed of which is the Census Block Group, a statistical division of the census tract, typically containing between 600 and 3,000 people.
Only the ACS 5-year estimates are published for Block Groups, which implies the 5-year data give the most detailed geographical resolution for demographic analysis.
ACS data tables are accessible in various ways, including the browser-based American FactFinder tool, FTP, and recently-introduced APIs.
The most comprehensive ACS data are contained in the ACS Summary File, a set of comma-delimited text files that contain all of the detailed tables for the ACS.
See the online Guidance for Data Users at https://www.census.gov/programs-surveys/acs/guidance.html for more information.
This project accesses and processes the ACS Summary File into detailed subject tables, according to the user's request for Year, State, and Detailed Table names.
The document How To construct ACS tables from Summary Files.docx included with this repo provides explanation for constructing detailed tables from the ACS Summary File.
The following excerpt from How To construct ACS tables from Summary Files.docx illustrates the logical flow of the censusACS.py python code.
Step-by-step example for creating Table B01002, “Median Age By Sex,” for all Colorado Census Block Groups.
- From Appendix A in file ACS_2015_SF_5YR_Appendices.xls, find the row containing table B01002 in the “Table Number” column. Note: Occasionally, but rarely, a table spans multiple sequence files, so be sure to look for all rows for a given table number.
- Record the row information found in Step #1 — Table Number (B01002), Table Title (Median Age By Sex), Geography Restrictions (none), Summary File Sequence Number (0003), Summary File Starting and Ending Positions (100-102) — for later use.
- From geography file g20155co.csv in Summary File Colorado_Tracts_Block_Groups_Only.zip, filter for and save the 3,532 rows containing Block Group summary level value (150) in the third column.
- From the geography file template 2015_SFGeoFileTemplate.xls in 2015_5yr_Summary_FileTemplates.zip, use Row #2 for the descriptive column names to associate with the geo data collected in Step #3 above.
- Using the “Logical Record Number” and “Geographic Identifier” columns for the block group data from Step #3, form a new table with these columns only.
- Referencing the Summary File Sequence Number 0003 from Step #2, open files e20155co0003000.txt and m20155co0003000.txt from the Summary File archive Colorado_Tracts_Block_Groups_Only.zip.
- Use the logical record numbers from the table constructed in Step #5 to locate the corresponding rows (matching column 6) in the estimate and margins files opened in Step #6.
- For each row in the estimate and margins files, obtain survey data values from the columns between the Summary File Starting and Ending Positions (inclusive) obtained from Appendix A in Step #2.
- Use the appropriate template file (e.g. Seq3.xls) from file 2015_5yr_Summary_FileTemplates.zip to get descriptive column names to associate with the survey data collected in Step #8.
- Merge by Logical Record Number the tables created in Step #5 and Step #9, then form a final output table by retaining only the Geographic Identifier column, estimate columns, and margin-of-error columns.
- Save the output table to a CSV file or database.