pyerualjetwork 4.2.2b1__py3-none-any.whl → 4.2.2b2__py3-none-any.whl

pyerualjetwork/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "4.2.2b1"
+__version__ = "4.2.2b2"
 __update__ = "* Changes: https://github.com/HCB06/PyerualJetwork/blob/main/CHANGES\n* PyerualJetwork Homepage: https://github.com/HCB06/PyerualJetwork/tree/main\n* PyerualJetwork document: https://github.com/HCB06/PyerualJetwork/blob/main/Welcome_to_PyerualJetwork/PYERUALJETWORK_USER_MANUEL_AND_LEGAL_INFORMATION(EN).pdf\n* YouTube tutorials: https://www.youtube.com/@HasanCanBeydili"
 
 def print_version(__version__):

{pyerualjetwork-4.2.2b1.dist-info → pyerualjetwork-4.2.2b2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pyerualjetwork
-Version: 4.2.2b1
+Version: 4.2.2b2
 Summary: PyerualJetwork is a machine learning library supported with GPU(CUDA) acceleration written in Python for professionals and researchers including with PLAN algorithm, PLANEAT algorithm (genetic optimization). Also includes data pre-process and memory manegament
 Author: Hasan Can Beydili
 Author-email: tchasancan@gmail.com

{pyerualjetwork-4.2.2b1.dist-info → pyerualjetwork-4.2.2b2.dist-info}/RECORD

@@ -1,9 +1,8 @@
-pyerualjetwork/__init__.py,sha256=S17EvWZTRS5-syTAdq8gQNAfuzqHEMStWH5I6fI5KnQ,641
+pyerualjetwork/__init__.py,sha256=6hlVOBT4G0fitxve2dAJrEjxFTTZ5YhWnmVTuG-dmLs,641
 pyerualjetwork/activation_functions.py,sha256=WWOdMd5pI6ZKe-ieKCIsKAYPQODHuXYxx7tzhA5xjes,11767
 pyerualjetwork/activation_functions_cuda.py,sha256=KmXJ5Cdig46XAMYakXFPEOlxSxtFJjD21-i3nGtxPjE,11807
 pyerualjetwork/data_operations.py,sha256=pb5CqJ0Th6fCjTNMCtqQMiwH3KezTxAijacglsKUxmY,14730
 pyerualjetwork/data_operations_cuda.py,sha256=UpoJoFhIwTU4xg9dVuLAxLAT4CkRaGsxvtJG9j1xrNo,17629
-pyerualjetwork/deneme.py,sha256=6b_g1ZyJRhPtqH5D5yUGeOKO6eHMwklUQenccs4G060,7869
 pyerualjetwork/help.py,sha256=nQ_YbYA2RtuafhuvkreNpX0WWL1I_nzlelwCtvei0_Y,775
 pyerualjetwork/loss_functions.py,sha256=6PyBI232SQRGuFnG3LDGvnv_PUdWzT2_2mUODJiejGI,618
 pyerualjetwork/loss_functions_cuda.py,sha256=C93IZJcrOpT6HMK9x1O4AHJWXYTkN5WZiqdssPbvAPk,617
@@ -19,7 +18,7 @@ pyerualjetwork/planeat_cuda.py,sha256=9uopmM-gTZpSb0EOExrOZPT8FF5BqDdEfCX0zYQb9Q
 pyerualjetwork/ui.py,sha256=wu2BhU1k-w3Kcho5Jtq4SEKe68ftaUeRGneUOSCVDjU,575
 pyerualjetwork/visualizations.py,sha256=QaYSIyVkJZ8NqpBKArQKkI1y37nCQo_KIM98IMssnRc,28766
 pyerualjetwork/visualizations_cuda.py,sha256=F60vQ92AXlMgBka3InXnOtGoM25vQJAlBIU2AlYTwks,29200
-pyerualjetwork-4.2.2b1.dist-info/METADATA,sha256=XBf-qok9_lB70pHC8EgwyPGxmCC9l-Penu0MiYc9Wvo,7914
-pyerualjetwork-4.2.2b1.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
-pyerualjetwork-4.2.2b1.dist-info/top_level.txt,sha256=BRyt62U_r3ZmJpj-wXNOoA345Bzamrj6RbaWsyW4tRg,15
-pyerualjetwork-4.2.2b1.dist-info/RECORD,,
+pyerualjetwork-4.2.2b2.dist-info/METADATA,sha256=Ay6UXzegPQKTwT7Eayo75DMGZg46fA2Qozd5-vW16AQ,7914
+pyerualjetwork-4.2.2b2.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
+pyerualjetwork-4.2.2b2.dist-info/top_level.txt,sha256=BRyt62U_r3ZmJpj-wXNOoA345Bzamrj6RbaWsyW4tRg,15
+pyerualjetwork-4.2.2b2.dist-info/RECORD,,

pyerualjetwork/deneme.py

@@ -1,171 +0,0 @@
-# -*- coding: utf-8 -*-
-"""
-Created on Thu Jun 20 03:55:15 2024
-
-@author: hasan
-"""
-
-import numpy as np
-from colorama import Fore
-from pyerualjetwork import plan, planeat, data_operations, model_operations
-from sklearn.linear_model import LogisticRegression
-from sklearn.ensemble import RandomForestClassifier
-import xgboost as xgb
-import tensorflow as tf
-from tensorflow.keras.models import Sequential
-from tensorflow.keras.layers import Dense
-from tensorflow.keras.optimizers import RMSprop
-from sklearn.metrics import classification_report, accuracy_score
-from sklearn.datasets import make_blobs
-from matplotlib import pyplot as plt
-
-fig, ax = plt.subplots(2, 3)  # Create a new figure and axe
-
-# Karar sınırı çizimi
-def plot_decision_boundary(x, y, model, feature_indices=[0, 1], h=0.02, model_name='str', ax=None, which_ax1=None, which_ax2=None, W=None, activation_potentiation=None):
-    """
-    Plot decision boundary by focusing on specific feature indices.
-    
-    Parameters:
-    - x: Input data
-    - y: Target labels
-    - model: Trained model
-    - feature_indices: Indices of the features to plot (default: [0, 1])
-    - h: Step size for the mesh grid
-    """
-    x_min, x_max = x[:, feature_indices[0]].min() - 1, x[:, feature_indices[0]].max() + 1
-    y_min, y_max = x[:, feature_indices[1]].min() - 1, x[:, feature_indices[1]].max() + 1
-    xx, yy = np.meshgrid(np.arange(x_min, x_max, h),
-                         np.arange(y_min, y_max, h))
-
-    grid = np.c_[xx.ravel(), yy.ravel()]
-    
-    # Create a full grid with zeros for non-selected features
-    grid_full = np.zeros((grid.shape[0], x.shape[1]))
-    grid_full[:, feature_indices] = grid
-    
-    if model == 'PLAN':
-        
-        Z = [None] * len(grid_full)
-
-        for i in range(len(grid_full)):
-            Z[i] = np.argmax(model_operations.predict_model_ram(grid_full[i], W=W, activation_potentiation=activation_potentiation))
-
-        Z = np.array(Z)
-        Z = Z.reshape(xx.shape)
-
-    else:
-
-        # Predict on the grid
-        #for i in range(len(grid_full)): grid_full[i] = activation_functions.apply_activation(grid_full[i], ['tanh', 'relu', 'elu', 'selu'])
-        Z = model.predict(grid_full)
-
-        if model_name == 'Deep Learning':
-            Z = np.argmax(Z, axis=1)  # Get class predictions
-
-        Z = Z.reshape(xx.shape)
-
-    # Plot decision boundary
-    ax[which_ax1, which_ax2].contourf(xx, yy, Z, alpha=0.8)
-    ax[which_ax1, which_ax2].scatter(x[:, feature_indices[0]], x[:, feature_indices[1]], c=np.argmax(y, axis=1), edgecolors='k', marker='o', s=20, alpha=0.9)
-    ax[which_ax1, which_ax2].set_xlabel(f'Feature {feature_indices[0] + 1}')
-    ax[which_ax1, which_ax2].set_ylabel(f'Feature {feature_indices[1] + 1}')
-    ax[which_ax1, which_ax2].set_title(model_name)
-
-
-X, y = make_blobs(n_samples=1000, centers=5, random_state=42)
-
-# Eğitim, test ve doğrulama verilerini ayırma
-x_train, x_test, y_train, y_test = data_operations.split(X, y, test_size=0.4, random_state=42) # For less train data use this: (X, y, test_size=0.9, random_state=42)
-
-# One-hot encoding işlemi
-y_train, y_test = data_operations.encode_one_hot(y_train, y_test)
-
-
-# Veri dengesizliği durumunu otomatik dengeleme
-x_train, y_train = data_operations.auto_balancer(x_train, y_train)
-scaler_params, x_train, x_test = data_operations.standard_scaler(x_train, x_test)
-
-# Lojistik Regresyon Modeli
-print(Fore.YELLOW + "------Lojistik Regresyon Sonuçları------" + Fore.RESET)
-lr_model = LogisticRegression(max_iter=1000, random_state=42)
-y_train_decoded = data_operations.decode_one_hot(y_train)
-lr_model.fit(x_train, y_train_decoded)
-
-y_test_decoded = plan.decode_one_hot(y_test)
-y_pred_lr = lr_model.predict(x_test)
-test_acc_lr = accuracy_score(y_test_decoded, y_pred_lr)
-print(f"Lojistik Regresyon Test Accuracy: {test_acc_lr:.4f}")
-print(classification_report(y_test_decoded, y_pred_lr))
-# Karar sınırını görselleştir
-plot_decision_boundary(x_test, y_test, lr_model, feature_indices=[0, 1], model_name='Logistic Regression', ax=ax, which_ax1=0, which_ax2=0)
-
-# Random Forest Modeli
-print(Fore.CYAN + "------Random Forest Sonuçları------" + Fore.RESET)
-rf_model = RandomForestClassifier(n_estimators=100, random_state=42)
-rf_model.fit(x_train, y_train_decoded)
-y_pred_rf = rf_model.predict(x_test)
-test_acc_rf = accuracy_score(y_test_decoded, y_pred_rf)
-print(f"Random Forest Test Accuracy: {test_acc_rf:.4f}")
-print(classification_report(y_test_decoded, y_pred_rf))
-# Karar sınırını görselleştir
-plot_decision_boundary(x_test, y_test, rf_model, feature_indices=[0, 1], model_name='Random Forest', ax=ax, which_ax1=0, which_ax2=1)
-
-# XGBoost Modeli
-print(Fore.MAGENTA + "------XGBoost Sonuçları------" + Fore.RESET)
-xgb_model = xgb.XGBClassifier(use_label_encoder=False, eval_metric='mlogloss', random_state=42)
-xgb_model.fit(x_train, y_train_decoded)
-y_pred_xgb = xgb_model.predict(x_test)
-test_acc_xgb = accuracy_score(y_test_decoded, y_pred_xgb)
-print(f"XGBoost Test Accuracy: {test_acc_xgb:.4f}")
-print(classification_report(y_test_decoded, y_pred_xgb))
-# Karar sınırını görselleştir
-plot_decision_boundary(x_test, y_test, xgb_model, feature_indices=[0, 1], model_name='XGBoost', ax=ax, which_ax1=0, which_ax2=2)
-
-# Derin Öğrenme Modeli (Yapay Sinir Ağı)
-
-input_dim = x_train.shape[1]  # Giriş boyutu
-
-
-model = Sequential()
-model.add(Dense(32, activation='relu', input_dim=input_dim))
-model.add(Dense(64, activation='tanh'))
-model.add(Dense(32, activation='relu'))
-model.add(Dense(64, activation='tanh'))
-model.add(Dense(y_train.shape[1], activation='softmax'))
-
-# Model derlemesi
-model.compile(optimizer=RMSprop(), loss='binary_crossentropy', metrics=['accuracy'])
-
-# Model eğitimi (early stopping ile)
-early_stop = tf.keras.callbacks.EarlyStopping(monitor='loss', patience=10, restore_best_weights=True)
-model.fit(x_train, y_train, epochs=5, batch_size=32, callbacks=[early_stop], verbose=1)
-
-# Test verileri üzerinde modelin performansını değerlendirme
-y_pred_dl = model.predict(x_test)
-y_pred_dl_classes = np.argmax(y_pred_dl, axis=1)  # Tahmin edilen sınıflar
-y_test_decoded_dl = plan.decode_one_hot(y_test)
-print(Fore.BLUE + "------Derin Öğrenme (ANN) Sonuçları------" + Fore.RESET)
-test_acc_dl = accuracy_score(y_test_decoded_dl, y_pred_dl_classes)
-print(f"Derin Öğrenme Test Accuracy: {test_acc_dl:.4f}")
-print(classification_report(y_test_decoded_dl, y_pred_dl_classes))
-# Karar sınırını görselleştir
-plot_decision_boundary(x_test, y_test, model=model, feature_indices=[0, 1], model_name='Deep Learning', ax=ax, which_ax1=1, which_ax2=0)
-
-# PLAN Modeli
-# Configuring optimizer
-genetic_optimizer = lambda *args, **kwargs: planeat.evolver(*args, **kwargs)
-
-model = plan.learner(x_train, y_train, genetic_optimizer, batch_size=0.05, gen=15) # learner function = TFL(Test Feedback Learning). If test parameters not given then uses Train Feedback. More information: https://github.com/HCB06/pyerualjetwork/blob/main/Welcome_to_PLAN/PLAN.pdf
-
-W = model[model_operations.get_weights()]
-activation_potentiation = model[model_operations.get_act_pot()]
-
-test_model = plan.evaluate(x_test, y_test, W=W, activation_potentiation=activation_potentiation)
-test_acc_plan = test_model[model_operations.get_acc()]
-print(Fore.GREEN + "------PLAN Modeli Sonuçları------" + Fore.RESET)
-print(f"PLAN Test Accuracy: {test_acc_plan:.4f}")
-print(classification_report(plan.decode_one_hot(y_test), test_model[model_operations.get_preds()]))
-# Karar sınırını görselleştir
-plot_decision_boundary(x_test, y_test, model='PLAN', feature_indices=[0, 1], model_name='PLAN', ax=ax, which_ax1=1, which_ax2=1, W=W, activation_potentiation=activation_potentiation)
-plt.show()