get_gender.py

__copyright__ = "Copyright (C) 2023 Ali Mustapha"
__license__ = "GPL-3.0-or-later"

import tensorflow as tf
import numpy as np
import re 
import unicodedata
from utils import data_utils
from unidecode import unidecode

# from utils import self
class GenderPredictor:
    def __init__(self, model_path):
        self.model = self.load_model(model_path)
        
    def load_model(self, path):
        model = tf.keras.models.load_model(path)
        # Compile and train the model
        model.compile(
            loss=tf.keras.losses.categorical_crossentropy,
            optimizer=tf.keras.optimizers.Adam(),
            metrics=['accuracy']
        )
        return model

    def predict_gender(self, name):
        EMAIL_re = re.compile(r"^[^\s@]+@[^\s@]+$") 
        proba=100
        
        if EMAIL_re.match(name):
            prediction = 2
        else:
            
            translator = str.maketrans(r"-._\/+", "      ")

            name = name.translate(translator)
            name = data_utils.text_to_romanize(name)
            name=data_utils.remove_spaces_from_ends(name)
            if (len(name) < 3 or data_utils.is_most_common_char(name)) and data_utils.is_roman_language(name):
                prediction = 2 
            elif not data_utils.is_alpha(name):
                prediction = 2
            else:
                translator = str.maketrans("", "", "0123456789")
                name = name.translate(translator)
                name = name.split()[0] if len(name.split()[0]) > 2 else name
                
            try:
                predictions_proba = self.model.predict([name], verbose=0).astype('float')
                prediction,proba = self.get_label(predictions_proba)
                prediction_part = []
                if prediction == 2:
                    parts = name.split()
                    if len(parts) >1:
                        for part in parts:
                            prediction = self.predict_gender(part)
                            prediction_part.append(prediction)
                        prediction=data_utils.find_common_item(prediction_part)
            except Exception:
                prediction = 2
        return prediction,proba

    def get_label(self, predictions_proba):
        for index, row in enumerate(predictions_proba):
            proba=100
            if row[2] >= 0.1:
                max_index = 2
            else:
                max_index = np.argmax(row)
                if max_index == 2:
                    proba= 100
                else:
                    proba =int(row[max_index] * 100)
                
                
        return max_index,proba