Changes made for the 2023 & 2026 elections

README.md

@@ -13,5 +13,16 @@ conda activate socrse2023
 pip install -r requirements.txt
 
 # Run voting
-python ranked_vote.py --token_col="Please insert your voting token here" ~/Downloads/SocRSE\ Trustee\ elections\ 2023.csv tokens.txt "Please rank the candidates (displayed in random order)" $num_places
+python ranked_vote.py --token_col="Please insert your voting token here" ~/Downloads/SocRSE\ Trustee\ elections\ 2023.csv tokens.txt "Candidate ranking" 7
+```
+
+### Token mapping
+
+In the 2026 election multiple sets of tokens were generated accidentally.
+A new `--token_map` argument has been created for such situations.
+It allows you to provide a 2-column CSV file, with column headings 'old' and 'new', mapping between valid tokens for each voter.
+For instance, you can run:
+
+```bash
+python ranked_vote.py --token_col="Please insert your voting token here" --token_map token_map.csv SocRSE\ Trustee\ elections\ 2026.csv tokens.txt "Candidate ranking" 5
 ```

ranked_vote.py

@@ -26,13 +26,15 @@
 parser.add_argument("question", type=str, help="The question name on the form")
 parser.add_argument("seats", type=int, help="Number of seats to elect")
 parser.add_argument("--token_col", type=str, help="The column in the CSV containing the voting token", default="Voting Token")
+parser.add_argument("--token_map", type=Path, default=None,
+                    help="A two column CSV file mapping old to new tokens. Column headings must be 'old' and 'new'")
 args = parser.parse_args()
 
-votes = parse_google_form(args.ballots, token_col=args.token_col)
+votes = parse_google_form(args.ballots, token_col=args.token_col, token_map=args.token_map)
 valid_tokens = parse_tokens(args.tokens)
 valid_votes, invalid_votes = filter_valid(votes, valid_tokens)
 
-print(f"Running election for {args.seats} seats")
+print(f"Running election for {args.seats} seats with {len(valid_votes)} valid votes ({len(invalid_votes)} invalid)")
 
 result = run_stv(valid_votes, args.question, args.seats)
 print(result)

tests/map.csv

@@ -0,0 +1,4 @@
+old,new
+"jdjfghdj","new1"
+"ghghdgn","new2"
+"jmhmjm","new3"

tests/test_voting.py

@@ -30,3 +30,10 @@ def test_stv(simple_file):
 def test_simple(simple_file):
     res = count_votes_simple(simple_file["Resolutions"]["Resolution 1"])
     assert res == (2, 3, 2, 5)
+
+
+def test_token_map():
+    test_dir = Path(__file__).resolve().parent
+    votes = parse_google_form(test_dir / "votes.csv", "Token", test_dir / "map.csv")
+    assert (votes.index[:3] == ["new1", "new2", "new3"]).all()
+    assert votes.index[3] == "hfgdhdfg"

utils.py

@@ -1,5 +1,6 @@
 import re
 from pathlib import Path
+from typing import Optional
 
 import pandas as pd
 import pyrankvote as rv
@@ -22,8 +23,13 @@ def parse_tokens(token_file: Path) -> set[str]:
         return set(t.strip() for t in tokens.readlines())
 
 
-def parse_google_form(csv_file: Path, token_col: str) -> pd.DataFrame:
+def parse_google_form(csv_file: Path, token_col: str, token_map: Optional[Path]=None) -> pd.DataFrame:
     votes = pd.read_csv(csv_file, dtype=str, keep_default_na=False)
+    if token_map:
+        token_mapping = pd.read_csv(token_map, dtype=str)
+        assert (token_mapping.columns == ["old", "new"]).all()
+        token_mapping = token_mapping.set_index("old").to_dict()["new"]
+        votes[token_col] = votes[token_col].replace(token_mapping)
     votes = votes.drop_duplicates(subset=[token_col], keep="last")
     votes = votes.set_index(token_col).drop(columns=["Timestamp"])
     headers = []