script eval

eval.py

+import json
+import pandas as pd
+import argparse
+
+def load_dictionary():
+    dictionary = {}
+    df = pd.read_csv("sens.csv", index_col="Acronym").drop("Unnamed: 0", axis=1)
+    for row in df.itertuples():
+        token = row[1]
+        lem_def = row[3]
+        if token not in dictionary:
+            dictionary[token] = []
+        dictionary[token].append(lem_def)
+    return dictionary
+
+def load_annot(path):
+    with open(path, "r") as file:
+        return json.load(file)
+
+def count(annot):
+    count_dict = {}
+
+    for d in annot:
+        _,_,_,acr = d.values()
+        for d2 in acr:
+            tid,tk,_,_,deflem,_,pred = d2.values()
+            if not tk in count_dict:
+                count_dict[tk] = {
+                    "count":[0,0],
+                    "senses":{}
+                }
+            
+            count_dict[tk]["count"][1] += 1
+            
+            if pred == "...":
+                print(f"--> Prédiction manquante pour id token {tid} ({tk})")
+            elif pred == deflem:
+                count_dict[tk]["count"][0] += 1
+    
+    return count_dict
+
+def acc(count_dict):
+    prec = 0
+    rapp = 0
+    somme = 0
+    prm = {k:None for k in count_dict.keys()}
+
+    for k,v in count_dict.items():
+        count = v["count"][0]
+        total = v["count"][1]
+
+        prec += count
+        rapp += total - count
+        somme += total
+        prm[k] = {
+            "precision" : count/total,
+            "rappel" : 1-count/total
+        }
+    
+    return prec/somme, rapp/somme, prm
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Eval modèle")
+    parser.add_argument(
+        "json",
+        action="store",
+        nargs=1,
+        help="Fichier json complété (champ prédiction renseigné)"
+    )
+    parser.add_argument(
+        "output",
+        action="store",
+        nargs="?",
+        help="Output (format json), optionnel"
+    )
+    args = parser.parse_args()
+
+    annot = load_annot(args.json[0])
+
+    prec, rapp, prm = acc(count(annot))
+
+    print("Précision :", prec)
+
+    if args.output:
+        with open(args.output, "w") as file:
+            json.dump(
+                {
+                    "precision": prec,
+                    "rappel" : rapp,
+                    "tokens": prm
+                },
+                file,
+                indent=4
+            )
\ No newline at end of file