Tools for working with the Familinx dataset.

sql/: PostgreSQL scripts for importing the data from Familinx.

• create_db.psql: Creates the profiles table
• delete_db.psql: Deletes the profiles table
• import_db.psql: Imports the database over the network. My database is in a virtual machine, you probably don't need it
• create_index.psql: Creates an index over the Familinx table on birth/death year; slows down query to unusable; do not run
• delete_index.psql: Deletes the index
• export_ids_alive_during_1809.psql: Exports the IDs of everyone alive during 1809; the output is used by the filtering scripts
• import_profile_ids_bfsout.psql: imports the node IDs output by ReadNodes.java into a table.
• create_joined_table_bfsout.psql: Creates a new filtered table containing only the descendants that were output by ReadNodes.java and imported by the above script

process_output/: processes data exported from the Familinx database.

• pick_one_in_twenty.py, picklist_ratio.py: takes a text file, sorts it, then pseudorandomly picks a percentage of the lines. The first script hardcodes it to one in twenty; the second script is configurable
• ReadNodes.java: takes the Familinx relations-anon.txt list of nodes and a list of nodes, and prints out all the descendants of those nodes.
• countsperyear_to_table.py: formats the output of the psql scripts that dump the number of people alive per year.

generate_sql: scripts that generate PostgreSQL files for exporting data

• make_queries_per_year.py: outputs script that computes how many people are alive in any given year. Doesn't count currently alive people; used for 1800-1900
• make_queries_per_year_alive.py: same as above, but takes currently alive people into account. Used for 1900-2010
• bfsout_make_queries_per_year.py: combination of the two above scripts, used to process the one-in-twenty filtered data
• varying_bfsout_make_queries_per_year.py: same script as above, but with tweakable table name; used to process 1%, 10%, 15%, and 20% filtered data
• import_profileids_ratio.py: combines import_profile_ids_bfsout.psql and create_joined_table_bfsout.psql for the 1/10/15/20% alternate ratio BFS output

graphtest: a sample family tree for testing ReadNodes.java, and scripts to graph the result.

Remove the first line (the field names) from profiles-anon.txt before importing into Postgres.

My code is licensed under the MIT license.

Data in the output/ directory is derived from the Familinx dataset and follows their license.