selection.py

from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
import numpy as np # Import numpy

class ExampleSelector:
    def __init__(self, df):
        self.vectorizer = TfidfVectorizer() # Define TF-IDF vectorizer
        self.embeddings = self.vectorizer.fit_transform(df["text"].tolist())
        self.texts = df["text"].tolist()
        self.labels = df["label"].tolist()
        self.df = df

    def most_similar(self, query, k=1):
        q_emb = self.vectorizer.transform([query]) # Embed training text
        sims = cosine_similarity(q_emb, self.embeddings).flatten() # Performs cosine similatiry search
        idxs = np.argsort(sims)[::-1][:k]
        return [(self.texts[i], self.labels[i]) for i in idxs]