doc strings

Author: mwunderl

sample-apps/aws-panorama-sample/code/lambda_function.py

@@ -15,6 +15,7 @@
 class people_counter(panoramasdk.base):
 
     def interface(self):
+        """Defines the application's configurable settings, input type, and output type."""
         return {
             "parameters":
                 (
@@ -34,6 +35,7 @@ def interface(self):
         }
 
     def init(self, parameters, inputs, outputs):
+        """Initializes the application's attributes with parameters from the interface, and default values."""
         try:
             self.threshold = parameters.threshold
             self.person_index = parameters.person_index
@@ -67,6 +69,7 @@ def init(self, parameters, inputs, outputs):
             return False
 
     def entry(self, inputs, outputs):
+        """Processes one frame of video from one or more video streams."""
         frame_start = time.time()
         self.frame_num += 1
         # Loop through attached video streams
@@ -92,6 +95,7 @@ def entry(self, inputs, outputs):
         return True
 
     def process_media(self, media):
+        """Runs inference on a buffered frame of video, and buffers the new frame."""
         stream = media.stream_uri
         # Set up stream buffer
         if not self.buffered_media.get(stream):
@@ -120,6 +124,7 @@ def process_media(self, media):
         return output
 
     def process_results(self, batch_set, output_media):
+        """Processes output tensors from a computer vision model and annotates a video frame."""
         # Model outputs (classes, probabilities, bounding boxes) are collected in
         # the BatchSet returned by model.get_result
         # Each output is a Batch of arrays, one for each input in the batch
@@ -147,6 +152,7 @@ def process_results(self, batch_set, output_media):
         return output_media
 
     def preprocess(self, img):
+        """Resizes and normalizes a frame of video."""
         resized = cv2.resize(img, (HEIGHT, WIDTH))
         mean = [0.485, 0.456, 0.406]
         std = [0.229, 0.224, 0.225]

sample-apps/custom-model/code/application.py

@@ -16,6 +16,7 @@
 class people_counter(panoramasdk.base):
 
     def interface(self):
+        """Defines the application's configurable settings, input type, and output type."""
         return {
             "parameters":
                 (
@@ -34,7 +35,7 @@ def interface(self):
         }
 
     def init(self, parameters, inputs, outputs):
-        """Load the model and allocate input and output arrays"""
+        """Initializes the application's attributes with parameters from the interface, and default values."""
         try:
             self.threshold = parameters.threshold
             self.person_index = parameters.person_index
@@ -64,7 +65,7 @@ def init(self, parameters, inputs, outputs):
             return False
 
     def entry(self, inputs, outputs):
-        """Process a frame of video from each input stream"""
+        """Processes one frame of video from one or more video streams."""
         frame_start = time.time()
         self.frame_num += 1
         # Loop through attached video streams
@@ -92,7 +93,7 @@ def entry(self, inputs, outputs):
         return True
 
     def process_media(self, media):
-        """Buffer image and run inference"""
+        """Runs inference on a buffered frame of video, and buffers the new frame."""
         stream = media.stream_uri
         # Set up stream buffer
         if not self.buffered_media.get(stream):
@@ -121,7 +122,7 @@ def process_media(self, media):
         return output
 
     def process_results(self, batch_set, output_media):
-        """Write text overlay on output image"""
+        """Processes output tensors from a computer vision model and annotates a video frame."""
         # Model outputs (classes, probabilities, bounding boxes) are collected in
         # the BatchSet returned by model.get_result
         # Each output is a Batch of arrays, one for each input in the batch
@@ -137,7 +138,7 @@ def process_results(self, batch_set, output_media):
         return output_media
 
     def preprocess(self, img):
-        """Resize and normalize input image"""
+        """Resizes and normalizes a frame of video."""
         resized = cv2.resize(img, (HEIGHT, WIDTH))
         mean = [0.485, 0.456, 0.406]
         std = [0.229, 0.224, 0.225]

sample-apps/custom-model/code/keras-model.py

@@ -20,7 +20,7 @@
 class KerasModel():
 
     def __init__(self):
-        """Get service role and bucket name from CloudFormation stack outputs"""
+        """Gets the service role and bucket name from CloudFormation stack outputs."""
         stack = cfn_client.describe_stacks(
             StackName='panorama-custom-model'
         )
@@ -31,7 +31,7 @@ def __init__(self):
         logger.info('Bucket: {}'.format(self.BUCKET_NAME))
 
     def compile(self, model_name='DenseNet121'):
-        """Compile a Keras application model"""
+        """Compiles a Keras application model."""
         #model_names = [model[0] for model in getmembers(tf.keras.applications, isfunction)]
         compilation_jobs = {}
         packaging_jobs = {}

sample-apps/custom-model/code/model.test.py

@@ -15,6 +15,7 @@
 class TestFunction(unittest.TestCase):
 
     def setUp(self):
+        """Gets resource names and IDs from the CloudFormation stack."""
         stack = cfn_client.describe_stacks(
             StackName='panorama-custom-model'
         )
@@ -23,6 +24,7 @@ def setUp(self):
         self.BUCKET_NAME=resources['bucketName']
 
     def test_function(self):
+        """Tests the compilation and packaging workflows with a single model."""
         #model_names = [model[0] for model in getmembers(tf.keras.applications, isfunction)]
         #model_names = ['ResNet50V2']
         #model_names = ['MobileNetV2']

sample-apps/custom-model/code/model/model.py

@@ -24,12 +24,12 @@
 class Model:
 
     def __init__(self, bucket_name, service_role_arn):
-        """Save bucket name and service role ARN"""
+        """Saves the bucket name and service role ARN."""
         self.bucket_name = bucket_name
         self.service_role_arn = service_role_arn
 
     def export_model(self, model_class, model_name):
-        """Export a TensorFlow model in SavedModel format and upload it to Amazon S3"""
+        """Exports a TensorFlow model in SavedModel format and uploads it to Amazon S3."""
         # TensorFlow 2 only
         # MODEL_NAME=model_class.__name__
         MODEL_NAME=model_name
@@ -55,7 +55,7 @@ def export_model(self, model_class, model_name):
         return model_uri, model_input_name, model_input_shape
 
     def compile_model(self, s3_uri, input_name='input_1', input_shape='224,224,3', framework='TENSORFLOW', device='jetson_xavier'):
-        """Compile a model with Amazon Sagemaker Neo"""
+        """Compiles a model with Amazon Sagemaker Neo."""
         MODEL_INPUT_SHAPE = '{{"{}":[1,{}]}}'.format(input_name,input_shape)
         COMPILED_MODEL_FOLDER_URI = 's3://{}/models-compiled'.format(self.bucket_name)
         COMPILATION_JOB = 'panorama-custom-model-'+ str(time.time()).split('.')[0]
@@ -80,7 +80,7 @@ def compile_model(self, s3_uri, input_name='input_1', input_shape='224,224,3', f
         return COMPILATION_JOB
 
     def package_model(self, model_name, compilation_job):
-        """Package a model with Amazon Sagemaker Neo"""
+        """Packages a model with Amazon Sagemaker Neo."""
         PACKAGING_JOB=compilation_job+"-packaging"
         MODEL_VERSION = "1.0"
         PACKAGED_MODEL_FOLDER_URI = 's3://{}/models-packaged'.format(self.bucket_name)
@@ -99,7 +99,7 @@ def package_model(self, model_name, compilation_job):
         return PACKAGING_JOB
 
     def wait_compilation(self, job):
-        """Wait for a Sagemaker Neo compilation job to complete"""
+        """Waits for a Sagemaker Neo compilation job to complete."""
         while True:
             response = sagemaker_client.describe_compilation_job(CompilationJobName=job)
             if response['CompilationJobStatus'] == 'COMPLETED':
@@ -110,7 +110,7 @@ def wait_compilation(self, job):
             time.sleep(5)
 
     def wait_packaging(self, job):
-        """Wait for a Sagemaker Neo packaging job to complete"""
+        """Waits for a Sagemaker Neo packaging job to complete."""
         while True:
             response = sagemaker_client.describe_edge_packaging_job(EdgePackagingJobName=job)
             if response['EdgePackagingJobStatus'] == 'COMPLETED':
@@ -121,13 +121,13 @@ def wait_packaging(self, job):
             time.sleep(5)
 
     def upload(self, bucket, key, src):
-        """Upload a file to Amazon S3"""
+        """Uploads a file to Amazon S3."""
         uri = 's3://{}/{}'.format(bucket,key)
         s3_resource.Bucket(bucket).Object(key).upload_file(src)
         return uri
 
     def remove(self, path):
-        """Remove a file or directory from /tmp"""
+        """Removes a file or directory from /tmp."""
         if os.path.exists(path) and path.startswith('/tmp/') and len(path) > 5:
             if os.path.isfile(path) or os.path.islink(path):
                 os.unlink(path)