replace logging.info with print

add pypi version in README.md
bug fix with x_only parameter for data_processor.py

README.md

@@ -1,5 +1,7 @@
 # ACES (Agricultural Classification and Estimation Service)
 
+[![image](https://img.shields.io/pypi/v/servir-aces.svg)](https://pypi.python.org/pypi/servir-aces)
+
 ACES (Agricultural Classification and Estimation Service) is a Python module for generating training data and training machine learning models for remote sensing applications. It provides functionalities for data processing, data loading from Earth Engine, feature extraction, and model training.
 
 ## Features

aces/data_processor.py

@@ -6,9 +6,6 @@
 This module provides functions for data input/output and preprocessing for the ACES project.
 """
 
-import logging
-logging.basicConfig(level=logging.INFO)
-
 import numpy as np
 import tensorflow as tf
 from functools import partial
@@ -81,7 +78,6 @@ def calculate_n_samples(**config):
                          patch_size=config.get("PATCH_SHAPE_SINGLE"),
                          features=config.get("FEATURES"),
                          labels=config.get("LABELS"),
-                         x_only=False,
                          depth=config.get("OUT_CLASS_NUM"))
 
         tf_training_records = tf.data.Dataset.list_files(f"{str(config.get('TRAINING_DIR'))}/*")\
@@ -117,19 +113,19 @@ def print_dataset_info(dataset: tf.data.Dataset, dataset_name: str) -> None:
 
         * dataset_name (str): The name of the dataset.
         """
-        logging.info(dataset_name)
+        print(dataset_name)
         for inputs, outputs in dataset.take(1):
             try:
-                logging.info(f"inputs: {inputs.dtype.name} {inputs.shape}")
-                logging.info(inputs)
-                logging.info(f"outputs: {outputs.dtype.name} {outputs.shape}")
-                logging.info(outputs)
+                print(f"inputs: {inputs.dtype.name} {inputs.shape}")
+                print(inputs)
+                print(f"outputs: {outputs.dtype.name} {outputs.shape}")
+                print(outputs)
             except:
-                logging.info(f" > inputs:")
+                print(f" > inputs:")
                 for name, values in inputs.items():
-                    logging.info(f"    {name}: {values.dtype.name} {values.shape}")
-                # logging.info(f"    example \n: {dataset.take(1)}")
-                logging.info(f" > outputs: {outputs.dtype.name} {outputs.shape}")
+                    print(f"    {name}: {values.dtype.name} {values.shape}")
+                # print(f"    example \n: {dataset.take(1)}")
+                print(f" > outputs: {outputs.dtype.name} {outputs.shape}")
 
     @staticmethod
     @tf.function
@@ -516,7 +512,7 @@ def get_dataset(pattern: str, features: list, labels: list, patch_size: int, bat
         Returns:
         * tf.data.Dataset: The TFRecord dataset.
         """
-        logging.info(f"Loading dataset from {pattern}")
+        print(f"Loading dataset from {pattern}")
 
         dataset = tf.data.Dataset.list_files(pattern).interleave(DataProcessor.create_tfrecord_from_file)
 
@@ -552,7 +548,7 @@ def get_dataset(pattern: str, features: list, labels: list, patch_size: int, bat
         # dataset = dataset.batch(batch_size)
         dataset = dataset.prefetch(buffer_size=tf.data.AUTOTUNE)
         if kwargs.get("training", False) and kwargs.get("TRANSFORM_DATA", True):
-            logging.info("randomly transforming data")
+            print("randomly transforming data")
             if kwargs.get("USE_AI_PLATFORM", False):
                 dataset = dataset.map(RandomTransform(), num_parallel_calls=tf.data.AUTOTUNE)
             else:

aces/ee_utils.py

@@ -1,8 +1,5 @@
 # -*- coding: utf-8 -*-
 
-import logging
-logging.basicConfig(level=logging.INFO)
-
 import re, warnings
 import ee
 import numpy as np
@@ -105,7 +102,7 @@ def export_collection_data(collection: ee.FeatureCollection, export_type: Union[
 
     @staticmethod
     def _export_collection_to_drive(collection, start_training, **kwargs) -> None:
-        logging.info("Exporting training data to Google Drive..")
+        print("Exporting training data to Google Drive..")
         training_task = ee.batch.Export.table.toDrive(
             collection=collection,
             description=kwargs.get("description", "myExportTableTask"),
@@ -119,7 +116,7 @@ def _export_collection_to_drive(collection, start_training, **kwargs) -> None:
     @staticmethod
     def _export_collection_to_asset(collection, start_training, **kwargs) -> None:
         asset_id = kwargs.get("asset_id", "myAssetId")
-        logging.info(f"Exporting training data to {asset_id}..")
+        print(f"Exporting training data to {asset_id}..")
         training_task = ee.batch.Export.table.toAsset(
             collection=collection,
             description=kwargs.get("description", "myExportTableTask"),
@@ -133,7 +130,7 @@ def _export_collection_to_cloud_storage(collection, start_training, **kwargs) ->
         description = kwargs.get("description", "myExportTableTask")
         bucket = kwargs.get("bucket", "myBucket")
         file_prefix = kwargs.get("file_prefix") if kwargs.get("file_prefix") is not None else description
-        logging.info(f"Exporting training data to gs://{bucket}/{file_prefix}..")
+        print(f"Exporting training data to gs://{bucket}/{file_prefix}..")
         training_task = ee.batch.Export.table.toCloudStorage(
             collection=collection,
             description=description,
@@ -157,7 +154,7 @@ def beam_export_collection_to_cloud_storage(collection_index, start_training, **
         bucket = kwargs.get("bucket", "myBucket")
         file_prefix = kwargs.get("file_prefix") if kwargs.get("file_prefix") is not None else description
         file_prefix = f"{file_prefix}_{index}"
-        logging.info(f"Exporting training data to gs://{bucket}/{file_prefix}..")
+        print(f"Exporting training data to gs://{bucket}/{file_prefix}..")
         training_task = ee.batch.Export.table.toCloudStorage(
             collection=collection,
             description=description,
@@ -197,7 +194,7 @@ def _export_image_to_cloud_storage(image, start_training, **kwargs) -> None:
             else:
                 raise ValueError(f"region must be an ee.FeatureCollection or ee.Geometry object. Found {type(region)}")
 
-        logging.info(f"Exporting training data to gs://{bucket}/{file_name_prefix}..")
+        print(f"Exporting training data to gs://{bucket}/{file_name_prefix}..")
 
         params = {
             "image": image,
@@ -224,7 +221,7 @@ def _export_image_to_cloud_storage(image, start_training, **kwargs) -> None:
     @staticmethod
     def _export_image_to_asset(image, start_training, **kwargs) -> None:
         asset_id = kwargs.get("asset_id", "")
-        logging.info(f"Exporting image to {asset_id}..")
+        print(f"Exporting image to {asset_id}..")
 
         training_task = ee.batch.Export.image.toAsset(
             image=image,

aces/metrics.py

@@ -8,9 +8,6 @@
 Additionally, it contains utility functions for plotting and visualizing model metrics during training.
 """
 
-import logging
-logging.basicConfig(level=logging.INFO)
-
 import matplotlib.pyplot as plt
 
 from tensorflow import keras

aces/model_builder.py

@@ -11,9 +11,6 @@
 provided specifications. The class also contains utility methods for constructing custom layers and defining metrics.
 """
 
-import logging
-logging.basicConfig(level=logging.INFO)
-
 import tensorflow as tf
 from tensorflow import keras
 
@@ -104,7 +101,7 @@ def build_and_compile_dnn_model(self, **kwargs):
             keras.Model: The compiled DNN model.
         """
         INITIAL_BIAS = kwargs.get("INITIAL_BIAS", None)
-        logging.info(f"INITIAL_BIAS: {INITIAL_BIAS}")
+        print(f"INITIAL_BIAS: {INITIAL_BIAS}")
 
         if INITIAL_BIAS is not None:
             INITIAL_BIAS = tf.keras.initializers.Constant(INITIAL_BIAS)
@@ -146,7 +143,7 @@ def build_and_compile_dnn_model_for_ai_platform(self, **kwargs):
             keras.Model: The compiled DNN model.
         """
         INITIAL_BIAS = kwargs.get("INITIAL_BIAS", None)
-        logging.info(f"INITIAL_BIAS: {INITIAL_BIAS}")
+        print(f"INITIAL_BIAS: {INITIAL_BIAS}")
         # DNN_DURING_ONLY = kwargs.get("DURING_ONLY", False)
 
         DERIVE_FEATURES = kwargs.get("DERIVE_FEATURES", False)
@@ -236,7 +233,7 @@ def build_and_compile_unet_model_for_ai_platform(self, **kwargs):
             with tf.distribute.MirroredStrategy().scope():
                 return self._build_and_compile_unet_model_for_ai_plaform(**kwargs)
         else:
-            logging.info("No distributed strategy found.")
+            print("No distributed strategy found.")
             return self._build_and_compile_unet_model(**kwargs)
 
     def _build_and_compile_unet_model_for_ai_plaform(self, **kwargs):
@@ -276,7 +273,7 @@ def build_and_compile_unet_model(self, **kwargs):
                 model.compile(optimizer=self.optimizer, loss=self.loss, metrics=metrics_list)
                 return model
         else:
-            logging.info("No distributed strategy found.")
+            print("No distributed strategy found.")
             model = self._build_and_compile_unet_model(**kwargs)
             metrics_list = [
                 Metrics.precision(),

aces/model_trainer.py

@@ -1,9 +1,5 @@
 # -*- coding: utf-8 -*-
 
-import logging
-logging.basicConfig(level=logging.INFO)
-# logger = logging.getLogger(__name__)
-
 import os
 import datetime
 import json
@@ -50,7 +46,7 @@ def __init__(self, config: Config, use_seed: bool = True, seed: int = 42):
         if self.use_seed:
             # producable results
             import random
-            logging.info(f"Using seed: {self.seed}")
+            print(f"Using seed: {self.seed}")
             tf.random.set_seed(self.seed)
             np.random.seed(self.seed)
             random.seed(2)
@@ -92,56 +88,56 @@ def train_model(self) -> None:
 
         7. Saves training parameters, plots, and models.
         """
-        logging.info("****************************************************************************")
-        logging.info("****************************** Clear Session... ****************************")
+        print("****************************************************************************")
+        print("****************************** Clear Session... ****************************")
         keras.backend.clear_session()
-        logging.info("****************************************************************************")
+        print("****************************************************************************")
         print(f"****************************** Configure memory growth... ************************")
         physical_devices = TFUtils.configure_memory_growth()
         self.config.physical_devices = physical_devices
-        logging.info("****************************************************************************")
-        logging.info("****************************** creating datasets... ************************")
+        print("****************************************************************************")
+        print("****************************** creating datasets... ************************")
         self.create_datasets(print_info=self.config.PRINT_INFO)
 
         if self.config.USE_AI_PLATFORM:
-            logging.info("****************************************************************************")
-            logging.info("******* building and compiling model for ai platform... ********************")
+            print("****************************************************************************")
+            print("******* building and compiling model for ai platform... ********************")
             self.build_and_compile_model_ai_platform()
         else:
-            logging.info("****************************************************************************")
-            logging.info("************************ building and compiling model... *******************")
+            print("****************************************************************************")
+            print("************************ building and compiling model... *******************")
             self.build_and_compile_model(print_model_summary=True)
 
-        logging.info("****************************************************************************")
-        logging.info("************************ preparing output directory... *********************")
+        print("****************************************************************************")
+        print("************************ preparing output directory... *********************")
         self.prepare_output_dir()
-        logging.info("****************************************************************************")
-        logging.info("****************************** training model... ***************************")
+        print("****************************************************************************")
+        print("****************************** training model... ***************************")
         self.start_training()
 
         if self.config.USE_AI_PLATFORM:
-            logging.info(self.model.summary())
+            print(self.model.summary())
 
-        logging.info("****************************************************************************")
-        logging.info("****************************** evaluating model... *************************")
+        print("****************************************************************************")
+        print("****************************** evaluating model... *************************")
         self.evaluate_and_print_val()
-        logging.info("****************************************************************************")
-        logging.info("****************************** saving parameters... ************************")
+        print("****************************************************************************")
+        print("****************************** saving parameters... ************************")
         ModelTrainer.save_parameters(**self.config.__dict__)
-        logging.info("****************************************************************************")
-        logging.info("*************** saving model config and history object... ******************")
+        print("****************************************************************************")
+        print("*************** saving model config and history object... ******************")
         self.save_history_object()
         if self.config.USE_AI_PLATFORM:
             ModelTrainer.save_model_config(self.config.MODEL_SAVE_DIR, **self._model.get_config())
         else:
             ModelTrainer.save_model_config(self.config.MODEL_SAVE_DIR, **self.model.get_config())
-        logging.info("****************************************************************************")
-        logging.info("****************************** saving plots... *****************************")
+        print("****************************************************************************")
+        print("****************************** saving plots... *****************************")
         self.save_plots()
-        logging.info("****************************************************************************")
-        logging.info("****************************** saving models... ****************************")
+        print("****************************************************************************")
+        print("****************************** saving models... ****************************")
         self.save_models()
-        logging.info("****************************************************************************")
+        print("****************************************************************************")
 
     def prepare_output_dir(self) -> None:
         """
@@ -151,7 +147,7 @@ def prepare_output_dir(self) -> None:
         """
         if not self.config.AUTO_MODEL_DIR_NAME:
             self.config.MODEL_SAVE_DIR = self.config.OUTPUT_DIR / self.config.MODEL_DIR_NAME
-            logging.info(f"> Saving models and results at {self.config.MODEL_SAVE_DIR}...")
+            print(f"> Saving models and results at {self.config.MODEL_SAVE_DIR}...")
             os.mkdir(self.config.MODEL_SAVE_DIR)
         else:
             today = datetime.date.today().strftime("%Y_%m_%d")
@@ -162,11 +158,11 @@ def prepare_output_dir(self) -> None:
                 try:
                     os.mkdir(self.config.MODEL_SAVE_DIR)
                 except FileExistsError:
-                    logging.info(f"> {self.config.MODEL_SAVE_DIR} exists, creating another version...")
+                    print(f"> {self.config.MODEL_SAVE_DIR} exists, creating another version...")
                     iterator += 1
                     continue
                 else:
-                    logging.info(f"> Saving models and results at {self.config.MODEL_SAVE_DIR}...")
+                    print(f"> Saving models and results at {self.config.MODEL_SAVE_DIR}...")
                     break
 
     def create_datasets(self, print_info: bool = False) -> None:
@@ -209,7 +205,7 @@ def create_datasets(self, print_info: bool = False) -> None:
         )
 
         if print_info:
-            logging.info("Printing dataset info:")
+            print("Printing dataset info:")
             DataProcessor.print_dataset_info(self.TRAINING_DATASET, "Training")
             DataProcessor.print_dataset_info(self.TESTING_DATASET, "Testing")
             DataProcessor.print_dataset_info(self.VALIDATION_DATASET, "Validation")
@@ -226,7 +222,7 @@ def build_and_compile_model(self, print_model_summary: bool = True) -> None:
         Prints the model summary if print_model_summary is set to True.
         """
         self.model = self.build_model(**self.config.__dict__)
-        if print_model_summary:  logging.info(self.model.summary())
+        if print_model_summary:  print(self.model.summary())
 
     def build_and_compile_model_ai_platform(self) -> None:
         """
@@ -240,7 +236,7 @@ def build_and_compile_model_ai_platform(self) -> None:
         Prints the model summary if print_model_summary is set to True.
         """
         model, wrapped_model = self.build_model(**self.config.__dict__)
-        logging.info(model.summary())
+        print(model.summary())
         self._model = model
         self.model = wrapped_model
 
@@ -309,16 +305,16 @@ def evaluate_and_print_val(self) -> None:
 
         Evaluates the model on the validation dataset and prints the metrics.
         """
-        logging.info("************************************************")
-        logging.info("************************************************")
-        logging.info("Validation")
+        print("************************************************")
+        print("************************************************")
+        print("Validation")
         # Tip: You can remove steps=self.config.TEST_SIZE and match the TEST_SIZE from the env
         evaluate_results = self.model.evaluate(self.TESTING_DATASET) # , steps=self.config.TEST_SIZE
         with open(f"{self.config.MODEL_SAVE_DIR}/evaluation.txt", "w") as evaluate:
             evaluate.write(json.dumps(dict(zip(self.model.metrics_names, evaluate_results))))
         for name, value in zip(self.model.metrics_names, evaluate_results):
-            logging.info(f"{name}: {value}")
-        logging.info("\n")
+            print(f"{name}: {value}")
+        print("\n")
 
     @staticmethod
     def save_parameters(**config) -> None:
@@ -368,7 +364,7 @@ def save_plots(self) -> None:
 
         Saves the model architecture plot, training history plot, and model object.
         """
-        logging.info(f"Saving plots and model visualization at {self.config.MODEL_SAVE_DIR}...")
+        print(f"Saving plots and model visualization at {self.config.MODEL_SAVE_DIR}...")
         if self.config.USE_AI_PLATFORM:
             keras.utils.plot_model(self._model, f"{self.config.MODEL_SAVE_DIR}/model.png", show_shapes=True, rankdir="TB")
             keras.utils.plot_model(self.model, f"{self.config.MODEL_SAVE_DIR}/wrapped_model.png", show_shapes=True, rankdir="LR") # rankdir='TB'