Optional: Coding assistance

Writing code for simple tasks such as showing images or iterating over files in a folder can be time-consuming. Generative artificial intelligence tools such as github copilot or bia-bob can help out. You can install bia-bob in your conda environment by opening another terminal, activating your environment and using pip:

Activate your environment:

mamba activate devbio-napari-env

Install bia-bob:

pip install bia-bob openai

Afterwards, restart the kernel of this notebook.

In this notebook we will demonstrate how to use bia-bob, a coding assistant specialized in bio-image analysis.

## The following cells might be necessary if you run this code on your computer. 
## Ask a trainer for the key to enter at ... 
import os
os.environ['OPENAI_API_KEY'] = 'sk-...'

You can initialize bia-bob like this.

import bia_bob

Afterwards, you ask for assistance by writing %bob in front of an English prompt. Bob will generate code right below the cell.

%bob load blobs.tif and show it

I will load an image file named blobs.tif from the disk and display it.

from skimage.io import imread
import stackview

# Load the image from disk
filename = 'blobs.tif'
image = imread(filename)

# Display the image
stackview.insight(image)

If your task is more complex, you can write multi-line prompts behind ´%%bob´. The more details you provide, e.g. which libraries and functions Bob should use, will make the code fit better to what you expect.

%%bob
For all image files in the folder ´data_banana´:
* load the image
* show the image using ´skimage.io.imshow´

I will load all image files located in the folder data_banana and display each one using skimage.io.imshow.

import os
from skimage.io import imread, imshow
import matplotlib.pyplot as plt

# Get a list of files in 'data_banana' directory
directory_path = 'data_banana'
file_list = os.listdir(directory_path)

# Filter out files that are image files (common extensions)
image_files = [f for f in file_list if f.endswith(('.tif', '.czi', '.png', '.jpg'))]

# Load and display each image file
for filename in image_files:
    # Construct the full image path
    file_path = os.path.join(directory_path, filename)
    
    # Load the image
    img = imread(file_path)
    
    # Display the image
    imshow(img)
    plt.title(f"Displaying image: {filename}")
    plt.show()