Author: Neal McBurnett
Revision Date: 2010-11-10 17:47:11 UTC

2010-11-10 <neal@bcn.boulder.co.us>
 * Add Kaplan-Markov tabulation data report at /kmreports/<contest>/
 * Add km_report view and kmreport.html to generate it
 * Add km_select_units to select audit units for Kaplan-Markov
 * Add Kaplan-Markov selection report at /kmselections/<contest>/
 * Add km_selection_report and kmselectionreport.html to generate it
 * Add parse_boulder_csv to parse Boulder-style audit data and testdata/test-boulder.csv
 * Add tests: TestBoulderCsv for parse_boulder_csv and TestSWDB_Csv_KM for Kaplan-Markov
 * Add "margin_offset" to Contest model to allow adjusting for untabulated ballots
 * Add "replace" and "prng" options to erandom.weightedsample() and speed it up
 * Add another way to combine audit units for privacy: Batch.merge() and ContestBatch.merge()
 * Add simple worst-case ContestBatch.taintfactor() calculation method
 * Automatically recognize different kinds of csv in parser by peeking at header line
 * Fix parsing bug: use get_or_create in au.save() to avoid error when ballot counts differ
    in batch get_or_create: sqlite3.IntegrityError: columns name, election_id, type are not unique
 * Add timestamps to parser output
 * Add crude util.input_sub_tallies() function to enter sub-batch tallies and add them up by columns.
 * Update notes in ToDo