PyGamLab 1.6.0__tar.gz → 1.7.0__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (46) hide show
  1. {pygamlab-1.6.0 → pygamlab-1.7.0}/PKG-INFO +1 -1
  2. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/__init__.py +2 -2
  3. pygamlab-1.7.0/PyGamLab/ai_core/__init__.py +18 -0
  4. pygamlab-1.7.0/PyGamLab/ai_core/gam_ai.py +490 -0
  5. pygamlab-1.7.0/PyGamLab/databases/Aflow.py +447 -0
  6. pygamlab-1.7.0/PyGamLab/databases/COD.py +357 -0
  7. pygamlab-1.7.0/PyGamLab/databases/Jarvis.py +398 -0
  8. pygamlab-1.7.0/PyGamLab/databases/Main_DB.py +533 -0
  9. pygamlab-1.7.0/PyGamLab/databases/Material_projects.py +364 -0
  10. pygamlab-1.7.0/PyGamLab/databases/__init__.py +22 -0
  11. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/GAM_architectures/GAM_Graphene.py +34 -29
  12. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/gamvis/gamvis_engine.py +1 -1
  13. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab.egg-info/PKG-INFO +1 -1
  14. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab.egg-info/SOURCES.txt +8 -0
  15. {pygamlab-1.6.0 → pygamlab-1.7.0}/pyproject.toml +1 -1
  16. {pygamlab-1.6.0 → pygamlab-1.7.0}/setup.py +1 -1
  17. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/Constants.py +0 -0
  18. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/Converters.py +0 -0
  19. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/Data_Analysis.py +0 -0
  20. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/Functions.py +0 -0
  21. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/GAM_architectures/GAM_nano_particles.py +0 -0
  22. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/GAM_architectures/GAM_nanotubes.py +0 -0
  23. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/GAM_architectures/GAM_phosphorene.py +0 -0
  24. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/GAM_architectures/GAM_sillicene.py +0 -0
  25. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/GAM_architectures/__init__.py +0 -0
  26. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/Generators/__init__.py +0 -0
  27. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/Generators/auto_bulk.py +0 -0
  28. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/Generators/one_d.py +0 -0
  29. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/Generators/two_d.py +0 -0
  30. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/Generators/zero_d.py +0 -0
  31. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/Primatom/__init__.py +0 -0
  32. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/Primatom/gam_molecule.py +0 -0
  33. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/Primatom/gam_universe.py +0 -0
  34. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/__init__.py +0 -0
  35. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/gamvis/__init__.py +0 -0
  36. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/gamvis/gamvis.py +0 -0
  37. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/io/__init__.py +0 -0
  38. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/io/checker.py +0 -0
  39. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/io/conversions.py +0 -0
  40. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/io/export.py +0 -0
  41. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab/structures/io/read.py +0 -0
  42. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab.egg-info/dependency_links.txt +0 -0
  43. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab.egg-info/requires.txt +0 -0
  44. {pygamlab-1.6.0 → pygamlab-1.7.0}/PyGamLab.egg-info/top_level.txt +0 -0
  45. {pygamlab-1.6.0 → pygamlab-1.7.0}/README.md +0 -0
  46. {pygamlab-1.6.0 → pygamlab-1.7.0}/setup.cfg +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: PyGamLab
3
- Version: 1.6.0
3
+ Version: 1.7.0
4
4
  Summary: A scientific Python library with constants, unit convertors, formulas, and data analysis tools.
5
5
  Home-page: https://github.com/APMaii/pygamlab/tree/main
6
6
  Author: Ali Pilehvar Meibody
@@ -43,11 +43,11 @@ Y88b. 88 Y88..88PP. 888 888 888 888...... Y88..88PP. 888 888
43
43
  '''
44
44
 
45
45
 
46
- __version__ = "1.6.0"
46
+ __version__ = "1.7.0"
47
47
 
48
48
 
49
49
 
50
- __all__ = ["Constants", "Converters", "Functions", "Data_Analysis","structures","databases"]
50
+ __all__ = ["Constants", "Converters", "Functions", "Data_Analysis","structures","databases","ai_core"]
51
51
 
52
52
 
53
53
 
@@ -0,0 +1,18 @@
1
+
2
+ from .gam_ai import GamAI_io
3
+ from .gam_ai import Gam_Ai_Workflow
4
+
5
+
6
+
7
+
8
+ __all__ = ["GamAI_io", "Gam_Ai_Workflow"]
9
+
10
+
11
+
12
+
13
+
14
+
15
+
16
+
17
+
18
+
@@ -0,0 +1,490 @@
1
+ import json, base64, joblib, io, os
2
+
3
+ import numpy as np
4
+ import matplotlib.pyplot as plt
5
+ from sklearn.metrics import (
6
+ confusion_matrix, ConfusionMatrixDisplay,
7
+ classification_report, r2_score, mean_absolute_error, mean_squared_error
8
+ )
9
+
10
+ class GamAI_io:
11
+ """
12
+ A unified input/output handler for saving and loading `.gam_ai` model packages.
13
+
14
+ This class encapsulates both machine learning model serialization and
15
+ relevant metadata (e.g., author info, model parameters, training details)
16
+ into a single portable `.gam_ai` file. It enables seamless model deployment,
17
+ archival, and reproducibility by combining the binary model object and
18
+ human-readable metadata in one structured JSON container.
19
+
20
+ Parameters
21
+ ----------
22
+ model_name : str, optional
23
+ A short, descriptive name for the model. Used as the filename during saving.
24
+ model_type : str, optional
25
+ The algorithmic or architectural family of the model
26
+ (e.g., "RandomForest", "XGBoost", "CNN", "Transformer").
27
+ description : str, optional, default=""
28
+ A brief summary describing the model’s purpose, training dataset,
29
+ or key features.
30
+ author_name : str, optional, default=""
31
+ Full name of the model creator.
32
+ author_email : str, optional, default=""
33
+ Contact email for correspondence or citation.
34
+ trainer_name : str, optional, default=""
35
+ The individual or system responsible for model training.
36
+ best_accuracy : float, optional
37
+ The highest validation or test accuracy achieved during training.
38
+ doi : str, optional
39
+ Digital Object Identifier (DOI) associated with the published model or dataset.
40
+ hyperparam_range : dict, optional, default={}
41
+ Dictionary defining hyperparameter search ranges used during optimization.
42
+ best_params : dict, optional, default={}
43
+ Dictionary of the final optimized hyperparameter values.
44
+ ml_model : object, optional
45
+ The trained machine learning model instance (e.g., scikit-learn estimator).
46
+
47
+ Notes
48
+ -----
49
+ The `.gam_ai` file format consists of a JSON object with two main sections:
50
+
51
+ - **metadata**: Contains descriptive fields such as author, model type,
52
+ and hyperparameters.
53
+ - **model_data**: Contains the Base64-encoded binary serialization of the
54
+ trained model, produced via `joblib`.
55
+
56
+ This approach ensures full portability and JSON readability, enabling
57
+ both programmatic and manual inspection of model metadata.
58
+
59
+ Examples
60
+ --------
61
+ >>> from gamai_io import GamAI_io
62
+ >>> from sklearn.ensemble import RandomForestClassifier
63
+ >>> model = RandomForestClassifier(n_estimators=100, random_state=42)
64
+ >>> model.fit(X_train, y_train)
65
+ >>>
66
+ >>> package = GamAI_io(
67
+ ... model_name="rf_classifier_v1",
68
+ ... model_type="RandomForest",
69
+ ... description="Predicts material phases using compositional data",
70
+ ... author_name="John Doe",
71
+ ... author_email="john.doe@example.com",
72
+ ... best_accuracy=0.94,
73
+ ... hyperparam_range={"n_estimators": [50, 100, 200]},
74
+ ... best_params={"n_estimators": 100, "max_depth": 10},
75
+ ... ml_model=model
76
+ ... )
77
+ >>>
78
+ >>> # Save to file
79
+ >>> package.save(save_dir="models")
80
+ 💾 Saved ml_model package: models/rf_classifier_v1.gam_ai
81
+ >>>
82
+ >>> # Load from file
83
+ >>> loaded_package = GamAI_io.load("models/rf_classifier_v1.gam_ai")
84
+ >>> restored_model = loaded_package.ml_model
85
+ >>> restored_model.predict(X_test[:5])
86
+
87
+ See Also
88
+ --------
89
+ joblib.dump : Efficient serialization of Python objects.
90
+ json : Standard JSON encoder/decoder.
91
+ base64 : Encoding binary model data for safe JSON storage.
92
+
93
+ """
94
+
95
+ def __init__(self, **kwargs):
96
+ self.model_name = kwargs.get("model_name")
97
+ self.model_type = kwargs.get("model_type")
98
+ self.description = kwargs.get("description", "")
99
+ self.author_name = kwargs.get("author_name", "")
100
+ self.author_email = kwargs.get("author_email", "")
101
+ self.trainer_name = kwargs.get("trainer_name", "")
102
+ self.best_accuracy = kwargs.get("best_accuracy", None)
103
+ self.doi=kwargs.get("doi", None)
104
+ self.hyperparam_range = kwargs.get("hyperparam_range", {})
105
+ self.best_params = kwargs.get("best_params", {})
106
+ self.ml_model = kwargs.get("ml_model", None)
107
+
108
+ # ----------- Save ----------
109
+ def save(self, save_dir="models"):
110
+ """
111
+ Serialize and save the current model and metadata as a `.gam_ai` package.
112
+
113
+ Parameters
114
+ ----------
115
+ save_dir : str, optional, default="models"
116
+ The target directory to save the `.gam_ai` file.
117
+ The directory will be created if it does not exist.
118
+
119
+ Raises
120
+ ------
121
+ ValueError
122
+ If no model (`ml_model`) is attached to the current instance.
123
+
124
+ Notes
125
+ -----
126
+ The model is serialized via `joblib` and encoded with Base64 to
127
+ ensure JSON compatibility. The resulting file can be safely shared
128
+ or uploaded to repositories without binary corruption.
129
+
130
+ """
131
+ os.makedirs(save_dir, exist_ok=True)
132
+ filename = f"{self.model_name}.gam_ai"
133
+ filepath = os.path.join(save_dir, filename)
134
+
135
+ if self.ml_model is None:
136
+ raise ValueError("No ml_model attached to save inside .gam_ai.")
137
+
138
+ buffer = io.BytesIO()
139
+ joblib.dump(self.ml_model, buffer)
140
+ buffer.seek(0)
141
+ encoded_model = base64.b64encode(buffer.read()).decode("utf-8")
142
+
143
+ data = {
144
+ "metadata": {k: v for k, v in self.__dict__.items() if k != "ml_model"},
145
+ "model_data": encoded_model
146
+ }
147
+
148
+ with open(filepath, "w") as f:
149
+ json.dump(data, f, indent=4)
150
+ print(f"💾 Saved ml_model package: {filepath}")
151
+
152
+ # ----------- Load ----------
153
+ @staticmethod
154
+ def load(filepath):
155
+ """
156
+ Load a `.gam_ai` model package from disk.
157
+
158
+ Parameters
159
+ ----------
160
+ filepath : str
161
+ Full path to the `.gam_ai` file to be loaded.
162
+
163
+ Returns
164
+ -------
165
+ GamAI_io
166
+ A `GamAI_io` instance containing both metadata and the
167
+ deserialized machine learning model (`ml_model`).
168
+
169
+ Notes
170
+ -----
171
+ The loading process reverses the Base64 encoding and `joblib`
172
+ serialization to reconstruct the original model object.
173
+
174
+ """
175
+ with open(filepath, "r") as f:
176
+ data = json.load(f)
177
+ metadata = data["metadata"]
178
+ model_data = base64.b64decode(data["model_data"])
179
+ buffer = io.BytesIO(model_data)
180
+ model = joblib.load(buffer)
181
+ return GamAI_io(**metadata, model=model)
182
+
183
+
184
+
185
+
186
+ class Gam_Ai_Workflow:
187
+ """
188
+ Intelligent, type-aware workflow manager for `.gam_ai` model packages.
189
+
190
+ This class provides a unified interface for loading, evaluating, refitting,
191
+ and visualizing machine learning models saved in the `.gam_ai` format.
192
+ It automatically detects the model type (classifier, regressor, or
193
+ unsupervised) and routes the evaluation pipeline accordingly.
194
+
195
+ Parameters
196
+ ----------
197
+ model_name : str
198
+ The name (without extension) of the `.gam_ai` file to be loaded.
199
+ base_dir : str, optional, default="gam_models"
200
+ Directory containing saved `.gam_ai` model files.
201
+
202
+ Raises
203
+ ------
204
+ FileNotFoundError
205
+ If the specified `.gam_ai` file cannot be located within `base_dir`.
206
+
207
+ Notes
208
+ -----
209
+ This class depends on the `GamAI_io` handler for deserializing `.gam_ai` files.
210
+ Each `.gam_ai` file contains both model metadata and a serialized model object.
211
+ Once loaded, `Gam_Ai_Workflow` provides:
212
+ - Smart evaluation (`evaluate()`)
213
+ - Visualization (`visualize_unsupervised()`)
214
+ - Refit capabilities (`refit()`)
215
+ - Summaries (`summary()`)
216
+
217
+ Examples
218
+ --------
219
+ >>> workflow = Gam_Ai_Workflow("rf_classifier_v1", base_dir="models")
220
+ ✅ Loaded model 'rf_classifier_v1' (classifier) successfully.
221
+ >>>
222
+ >>> workflow.summary()
223
+ 📘 MODEL SUMMARY
224
+ model_name: rf_classifier_v1
225
+ model_type: classifier
226
+ author_name: John Doe
227
+ best_accuracy: 0.94
228
+ ...
229
+ >>>
230
+ >>> workflow.evaluate(X_test, y_test)
231
+ 🎯 Accuracy: 0.9470
232
+ 📊 Classification Report:
233
+ ...
234
+ """
235
+ def __init__(self, model_name, base_dir="gam_models"):
236
+ self.model_name = model_name
237
+ self.base_dir = base_dir
238
+ self.file_path = os.path.join(base_dir, f"{model_name}.gam_ai")
239
+
240
+ if not os.path.exists(self.file_path):
241
+ raise FileNotFoundError(f"❌ .gam_ai file not found: {self.file_path}")
242
+
243
+ self.gam = GamAI_io.load(self.file_path)
244
+ self.model = self.gam.model
245
+ self.model_type = self.gam.model_type.lower()
246
+
247
+ print(f"✅ Loaded model '{self.model_name}' ({self.model_type}) successfully.")
248
+
249
+ # ---------- General ----------
250
+ def summary(self):
251
+ """
252
+ Display a concise summary of the loaded model and its metadata.
253
+
254
+ Prints all metadata fields stored in the `.gam_ai` file,
255
+ including model type, author information, and hyperparameters.
256
+
257
+ Returns
258
+ -------
259
+ None
260
+ """
261
+ print("\n📘 MODEL SUMMARY")
262
+ for k, v in self.gam.__dict__.items():
263
+ if k != "model":
264
+ print(f"{k}: {v}")
265
+
266
+ def predict(self, X):
267
+ """
268
+ Run model inference on input data.
269
+
270
+ Parameters
271
+ ----------
272
+ X : array-like
273
+ Input features compatible with the trained model.
274
+
275
+ Returns
276
+ -------
277
+ np.ndarray
278
+ Model predictions corresponding to `X`.
279
+
280
+ Raises
281
+ ------
282
+ ValueError
283
+ If no model is loaded.
284
+ """
285
+ if self.model is None:
286
+ raise ValueError("Model not loaded.")
287
+ return self.model.predict(X)
288
+
289
+ def refit(self, X, y):
290
+ """
291
+ Retrain (refit) the loaded model on new data.
292
+
293
+ Parameters
294
+ ----------
295
+ X : array-like
296
+ Training features.
297
+ y : array-like
298
+ Corresponding training labels or targets.
299
+
300
+ Raises
301
+ ------
302
+ NotImplementedError
303
+ If the model does not support the `.fit()` method.
304
+
305
+ Notes
306
+ -----
307
+ This method modifies the current model in place and does not automatically
308
+ update the `.gam_ai` file on disk. To persist changes, re-save the model
309
+ using `GamAI_io.save()` after refitting.
310
+ """
311
+ if hasattr(self.model, "fit"):
312
+ self.model.fit(X, y)
313
+ print("🔁 Model refitted successfully.")
314
+ else:
315
+ raise NotImplementedError("This model type cannot be refitted.")
316
+
317
+ # ---------- Classifier ----------
318
+ def evaluate_classifier(self, X, y_true):
319
+ """
320
+ Evaluate a classification model and display key performance metrics.
321
+
322
+ Parameters
323
+ ----------
324
+ X : array-like
325
+ Input test features.
326
+ y_true : array-like
327
+ Ground-truth class labels.
328
+
329
+ Returns
330
+ -------
331
+ None
332
+
333
+ Notes
334
+ -----
335
+ This method computes and displays:
336
+ - Accuracy score
337
+ - Classification report
338
+ - Confusion matrix plot
339
+ """
340
+
341
+ y_pred = self.model.predict(X)
342
+ acc = np.mean(y_pred == y_true)
343
+ print(f"🎯 Accuracy: {acc:.4f}\n")
344
+ print("📊 Classification Report:\n")
345
+ print(classification_report(y_true, y_pred))
346
+
347
+ cm = confusion_matrix(y_true, y_pred)
348
+ disp = ConfusionMatrixDisplay(cm)
349
+ disp.plot(cmap="Blues")
350
+ plt.title(f"Confusion Matrix: {self.gam.model_name}")
351
+ plt.show()
352
+
353
+ # ---------- Regressor ----------
354
+ def evaluate_regressor(self, X, y_true):
355
+ """
356
+ Evaluate a regression model and visualize performance trends.
357
+
358
+ Parameters
359
+ ----------
360
+ X : array-like
361
+ Input test features.
362
+ y_true : array-like
363
+ True continuous target values.
364
+
365
+ Returns
366
+ -------
367
+ None
368
+
369
+ Notes
370
+ -----
371
+ This method prints and plots:
372
+ - Coefficient of determination (R²)
373
+ - Mean Absolute Error (MAE)
374
+ - Mean Squared Error (MSE)
375
+ - Predicted vs. True value scatter plot
376
+ """
377
+ y_pred = self.model.predict(X)
378
+ r2 = r2_score(y_true, y_pred)
379
+ mae = mean_absolute_error(y_true, y_pred)
380
+ mse = mean_squared_error(y_true, y_pred)
381
+
382
+ print(f"📈 R²: {r2:.4f}")
383
+ print(f"📉 MAE: {mae:.4f}")
384
+ print(f"📉 MSE: {mse:.4f}\n")
385
+
386
+ plt.figure(figsize=(6, 4))
387
+ plt.scatter(y_true, y_pred, alpha=0.6)
388
+ plt.plot([min(y_true), max(y_true)], [min(y_true), max(y_true)], 'r--')
389
+ plt.xlabel("True Values")
390
+ plt.ylabel("Predicted Values")
391
+ plt.title(f"Prediction Trend — {self.gam.model_name}")
392
+ plt.grid(True)
393
+ plt.show()
394
+
395
+ # ---------- Unsupervised ----------
396
+ def visualize_unsupervised(self, X):
397
+ """
398
+ Visualize cluster assignments or feature transformations for unsupervised models.
399
+
400
+ Parameters
401
+ ----------
402
+ X : array-like of shape (n_samples, n_features)
403
+ Input data to visualize.
404
+
405
+ Returns
406
+ -------
407
+ None
408
+
409
+ Raises
410
+ ------
411
+ NotImplementedError
412
+ If the model lacks both `predict()` and `transform()` methods.
413
+
414
+ Notes
415
+ -----
416
+ - If the model has a `.predict()` method, cluster assignments are plotted.
417
+ - If the model has a `.transform()` method, the transformed feature space is shown.
418
+ - This visualization assumes that the first two components or features
419
+ are suitable for 2D projection.
420
+ """
421
+ if hasattr(self.model, "predict"):
422
+ y_pred = self.model.predict(X)
423
+ plt.scatter(X[:, 0], X[:, 1], c=y_pred, cmap="viridis", s=30)
424
+ plt.title(f"Cluster Visualization — {self.gam.model_name}")
425
+ plt.show()
426
+ elif hasattr(self.model, "transform"):
427
+ X_trans = self.model.transform(X)
428
+ plt.scatter(X_trans[:, 0], X_trans[:, 1], s=30)
429
+ plt.title(f"Feature Space — {self.gam.model_name}")
430
+ plt.show()
431
+ else:
432
+ raise NotImplementedError("This unsupervised model has no visualization method.")
433
+
434
+ # ---------- Smart Evaluation ----------
435
+ def evaluate(self, X, y_true=None):
436
+ """
437
+ Automatically dispatch model evaluation based on its declared type.
438
+
439
+ Parameters
440
+ ----------
441
+ X : array-like
442
+ Input data.
443
+ y_true : array-like, optional
444
+ Ground-truth labels or target values (required for supervised models).
445
+
446
+ Raises
447
+ ------
448
+ ValueError
449
+ If the model type is not recognized.
450
+
451
+ Notes
452
+ -----
453
+ This method intelligently determines which evaluation routine to run:
454
+ - `evaluate_classifier()` for classification models
455
+ - `evaluate_regressor()` for regression models
456
+ - `visualize_unsupervised()` for unsupervised models
457
+
458
+ Examples
459
+ --------
460
+ >>> workflow.evaluate(X_test, y_test)
461
+ 🎯 Accuracy: 0.9470
462
+ 📊 Classification Report:
463
+ ...
464
+ """
465
+ if self.model_type == "classifier":
466
+ self.evaluate_classifier(X, y_true)
467
+ elif self.model_type == "regressor":
468
+ self.evaluate_regressor(X, y_true)
469
+ elif self.model_type == "unsupervised":
470
+ self.visualize_unsupervised(X)
471
+ else:
472
+ raise ValueError(f"Unknown model_type: {self.model_type}")
473
+
474
+
475
+
476
+
477
+
478
+
479
+
480
+
481
+
482
+
483
+
484
+
485
+
486
+
487
+
488
+
489
+
490
+