arehman10 · arehman10 · May 25, 2025
diff --git a/isic.csv b/isic.csv
@@ -0,0 +1,20 @@
+0111;Growing of cereals (except rice), leguminous crops and oil seeds
+0112;Growing of rice
+0113;Growing of vegetables and melons, roots and tubers
+0114;Growing of sugar cane
+0116;Growing of fibre crops
+0311;Marine fishing
+0510;Mining of hard coal
+1010;Processing and preserving of meat
+1071;Manufacture of bakery products
+1811;Printing
+2011;Manufacture of basic chemicals
+4610;Wholesale on a fee or contract basis
+4711;Retail sale in non-specialized stores with food, beverages or tobacco predominating
+5510;Short term accommodation activities
+5610;Restaurants and mobile food service activities
+6201;Computer programming activities
+7020;Management consultancy activities
+8610;Hospital activities
+9311;Operation of sports facilities
+9602;Hairdressing and other beauty treatment
diff --git a/isic_classifier.py b/isic_classifier.py
@@ -0,0 +1,57 @@
+import csv
+import re
+import sys
+from typing import List, Tuple
+
+
+def load_isic(path: str = "isic.csv") -> List[Tuple[str, str]]:
+    """Load ISIC code descriptions from a semicolon separated file."""
+    with open(path, newline='', encoding='utf-8') as f:
+        reader = csv.reader(f, delimiter=';')
+        return [(code.strip(), desc.strip()) for code, desc in reader]
+
+
+def _tokenize(text: str) -> set:
+    """Return a set of word tokens from the given text."""
+    return set(re.findall(r"\b\w+\b", text.lower()))
+
+
+def classify(activity: str, isic_data: List[Tuple[str, str]]) -> Tuple[str, float]:
+    """Return the ISIC code with the highest token overlap score."""
+    tokens = _tokenize(activity)
+    best_code = None
+    best_score = 0.0
+    for code, desc in isic_data:
+        desc_tokens = _tokenize(desc)
+        if not desc_tokens:
+            continue
+        score = len(tokens & desc_tokens) / len(desc_tokens)
+        if score > best_score:
+            best_score = score
+            best_code = code
+    return best_code, best_score
+
+
+def classify_file(input_csv: str, output_csv: str, column: str = "d1a1x", isic_path: str = "isic.csv") -> None:
+    """Classify activities in a CSV file and write results to a new CSV."""
+    isic_data = load_isic(isic_path)
+    with open(input_csv, newline='', encoding='utf-8') as f:
+        reader = csv.DictReader(f)
+        rows = list(reader)
+
+    for row in rows:
+        code, ratio = classify(row[column], isic_data)
+        row['isic_code'] = code
+        row['match_score'] = f"{ratio:.2f}"
+
+    with open(output_csv, 'w', newline='', encoding='utf-8') as f:
+        writer = csv.DictWriter(f, fieldnames=list(rows[0].keys()))
+        writer.writeheader()
+        writer.writerows(rows)
+
+
+if __name__ == "__main__":
+    if len(sys.argv) == 3:
+        classify_file(sys.argv[1], sys.argv[2])
+    else:
+        print("Usage: python isic_classifier.py input.csv output.csv")
diff --git a/sample_activities.csv b/sample_activities.csv
@@ -0,0 +1,6 @@
+d1a1x
+Growing of rice
+Computer software development
+Restaurants
+Short term accommodation services
+Operation of sports stadium
diff --git a/test_classifier.py b/test_classifier.py
@@ -0,0 +1,26 @@
+import csv
+import os
+from isic_classifier import classify_file
+
+
+def test_classify_file(tmp_path):
+    input_csv = os.path.join(tmp_path, 'input.csv')
+    output_csv = os.path.join(tmp_path, 'output.csv')
+    # copy sample activities file
+    with open('sample_activities.csv', 'r', encoding='utf-8') as src:
+        data = src.read()
+    with open(input_csv, 'w', encoding='utf-8') as dst:
+        dst.write(data)
+
+    classify_file(input_csv, output_csv)
+
+    with open(output_csv, newline='', encoding='utf-8') as f:
+        reader = csv.DictReader(f)
+        results = list(reader)
+
+    codes = [row['isic_code'] for row in results]
+    assert codes[0] == '0112'
+    assert codes[1] == '6201'
+    assert codes[2] == '5610'
+    assert codes[3] == '5510'
+    assert codes[4] == '9311'