Install draw.io Desktop on Linux Mint (should work for Debian as well): Download AppImage file from Github releases page, e.g. Release 27.0.9 · jgraph/drawio-desktop Make AppImage file executable: chmod u+x drawio.AppImage Move AppImage file to ~/.local/bin/ mv drawio.AppImage ~/.local/bin/drawio.AppImage Create and edit the desktop entry file nano ~/.local/share/applications/appimagekit-drawio.desktop Add following content to the desktop entry file and save changes [Desktop Entry] Name=draw.io Comment=Diagramming Application Exec=/home/YOUR_USERNAME/.local/bin/appimagekit-drawio.AppImage Icon=com.jgraph.drawio.desktop Terminal=false Type=Application Categories=Graphics; Update the desktop database xdg-desktop-menu install ~/.local/share/applications/appimagekit-drawio.desktop draw.io should appear in Linux Mint start menu under “Graphics”
Simple copy-paste augmentation for YOLO object detection task from PIL import Image import random def paste_image(source_path, target_path, position=None): """ Pastes a small source image onto a larger target image. Parameters: - source_path: Path to the small source image. - target_path: Path to the larger target image. - position: Optional; A tuple (x, y) specifying the top-left corner where the source image will be pasted. If not provided, a random position will be chosen. Returns: - new_image: The newly created image with the source image pasted onto it. - bbox_yolo: A tuple (x_center, y_center, width, height) representing the bounding box of the pasted image in YOLO format. """ # Open the source and target images source_image = Image.open(source_path) target_image = Image.open(target_path) # Get the dimensions of the source and target images source_width, source_height = source_image.size target_width, target_height = target_image.size # Choose a random position if not provided if position is None: max_x = target_width - source_width max_y = target_height - source_height if max_x < 0 or max_y < 0: raise ValueError("The source image is larger than the target image.") position = (random.randint(0, max_x), random.randint(0, max_y)) else: # Ensure the specified position is within bounds if position[0] < 0 or position[1] < 0 or position[0] + source_width > target_width or position[1] + source_height > target_height: raise ValueError("The specified position is out of the target image bounds.") # Paste the source image onto the target image target_image.paste(source_image, position) # Calculate the bounding box of the pasted image left = position[0] upper = position[1] right = position[0] + source_width lower = position[1] + source_height bbox = (left, upper, right, lower) # Convert the bounding box to YOLO format x_center = (left + right) / 2 / target_width y_center = (upper + lower) / 2 / target_height width = source_width / target_width height = source_height / target_height bbox_yolo = (x_center, y_center, width, height) return target_image, bbox, bbox_yolo # Example usage if __name__ == "__main__": source_path = "path/to/small_image.png" # Replace with the path to your small source image target_path = "path/to/large_image.png" # Replace with the path to your large target image # Example with a specified position new_image, bbox_yolo = paste_image(source_path, target_path, position=(50, 50)) new_image.save("path/to/new_image_specified_position.png") # Replace with the desired path to save the new image print("Bounding box of the pasted image (specified position) in YOLO format:", bbox_yolo) # Example with a random position new_image, bbox_yolo = paste_image(source_path, target_path) new_image.save("path/to/new_image_random_position.png") # Replace with the desired path to save the new image print("Bounding box of the pasted image (random position) in YOLO format:", bbox_yolo)
PySpark DataFrame - add leading zero if the value is a single digit between 0 and 9 from pyspark.sql import SparkSession from pyspark.sql.functions import col, when, regexp_extract, length, lpad # Create a SparkSession spark = SparkSession.builder.appName("TupleToDataFrame").getOrCreate() # List of tuples arr = [("a", "0a1x"), ("x", "3"), ("cc", "11"), ("h", "5")] # Create a DataFrame df = spark.createDataFrame(arr, ["column1", "column2"]) # Function to add leading zero if the value is a single digit between 0 and 9, # replace non-digit values with "0000", and leave other values as is df = df.withColumn("column2", when(regexp_extract(col("column2"), "^\\d+$", 0) == "", "0000") .when((regexp_extract(col("column2"), "^\\d+$", 0) != "") & (length(col("column2")) == 1), lpad(col("column2"), 2, "0")) .otherwise(col("column2")) ) # Show the updated DataFrame df.show()
Read files marked as deleted from an AWS S3 bucket into a DynamicFrame using boto3 and AWS Glue import boto3 from awsglue.context import GlueContext from pyspark.context import SparkContext from awsglue.dynamicframe import DynamicFrame def read_deleted_files_to_dynamicframe(bucket_name, prefix=''): # Initialize Glue context sc = SparkContext() glueContext = GlueContext(sc) # Initialize S3 client s3 = boto3.client('s3') # List object versions including delete markers paginator = s3.get_paginator('list_object_versions') page_iterator = paginator.paginate(Bucket=bucket_name, Prefix=prefix) # Collect keys of deleted files deleted_files = [] for page in page_iterator: if 'DeleteMarkers' in page: for delete_marker in page['DeleteMarkers']: if delete_marker['IsLatest']: deleted_files.append(delete_marker['Key']) # Read deleted files into a DynamicFrame if deleted_files: # Create a DynamicFrame from the deleted files dyf = glueContext.create_dynamic_frame.from_options( connection_type="s3", connection_options={ "paths": [f"s3://{bucket_name}/{key}" for key in deleted_files], "recurse": True, "useS3ListImplementation": True, "readVersion": "LATEST_PREVIOUS" # Read the version before the delete marker }, format="csv", # Adjust this based on your file format format_options={ "withHeader": True, "separator": "," } ) return dyf else: print("No deleted files found.") return None # Usage bucket_name = 'your-bucket-name' prefix = 'your-folder-prefix/' # Optional dyf = read_deleted_files_to_dynamicframe(bucket_name, prefix) if dyf: # Print the schema of the DynamicFrame dyf.printSchema() # Convert to DataFrame for further processing if needed df = dyf.toDF() # Perform your manipulations here # Convert back to DynamicFrame if necessary result_dyf = DynamicFrame.fromDF(df, glueContext, "result_dyf") # Write the result back to S3 glueContext.write_dynamic_frame.from_options( frame=result_dyf, connection_type="s3", connection_options={"path": "s3://output-bucket/output-path/"}, format="parquet" )
Read files from an AWS S3 bucket based on their creation date using AWS Glue import boto3 from datetime import datetime from awsglue.context import GlueContext from pyspark.context import SparkContext # First, use boto3 to list and filter objects def list_s3_files_after_date(bucket_name, prefix='', start_date=None): s3 = boto3.client('s3') paginator = s3.get_paginator('list_objects_v2') page_iterator = paginator.paginate(Bucket=bucket_name, Prefix=prefix) filtered_files = [] for page in page_iterator: if 'Contents' in page: for obj in page['Contents']: # S3 doesn't have a separate creation date, so we use LastModified if start_date is None or obj['LastModified'].replace(tzinfo=None) > start_date: filtered_files.append(obj['Key']) return filtered_files # Usage bucket_name = 'your-bucket-name' start_date = datetime(2024, 1, 1) # Files created after January 1, 2024 filtered_files = list_s3_files_after_date(bucket_name, start_date=start_date) # Then, use AWS Glue to create a DynamicFrame from the filtered files def read_filtered_files_to_dynamicframe(bucket_name, filtered_files): sc = SparkContext() glueContext = GlueContext(sc) if filtered_files: dyf = glueContext.create_dynamic_frame.from_options( connection_type="s3", connection_options={ "paths": [f"s3://{bucket_name}/{key}" for key in filtered_files], "recurse": True, "useS3ListImplementation": True }, format="csv", # Adjust this based on your file format format_options={ "withHeader": True, "separator": "," } ) return dyf else: print("No files found matching the criteria.") return None # Usage bucket_name = 'your-bucket-name' start_date = datetime(2024, 1, 1) # Files created after January 1, 2024 filtered_files = list_s3_files_after_date(bucket_name, start_date=start_date) dyf = read_filtered_files_to_dynamicframe(bucket_name, filtered_files) if dyf: # Print the schema of the DynamicFrame dyf.printSchema() # Convert to DataFrame for further processing if needed df = dyf.toDF() # Perform your manipulations here # Write the result back to S3 if needed glueContext.write_dynamic_frame.from_options( frame=dyf, connection_type="s3", connection_options={"path": "s3://output-bucket/output-path/"}, format="parquet" )