Use and Create Template

This section demonstrates how to create and use templates in the hsi-wizard framework for processing a DataCube. Templates help automate and reuse preprocessing workflows.

Create a Template

The following example shows how to create a processing template. We generate synthetic image data, apply a sequence of transformations (like resizing and background removal), and save the recorded operations into a template file.

import wizard
from wizard._utils.example import generate_pattern_stack
import matplotlib.pyplot as plt

# Prepare plot
fig, axes = plt.subplots(1, 2, figsize=(10, 8))

# Create synthetic data and initialize DataCube
data1 = generate_pattern_stack(20, 300, 300, n_circles=10, n_rects=0, n_triangles=0, seed=42)
dc1 = wizard.DataCube(data1)

# Plot before processing
axes[0].imshow(dc1.cube[10], cmap='viridis', aspect='auto')
axes[0].set_title("Before Processing DC1")

# Record and apply processing
dc1.start_recording()
dc1.resize(x_new=500, y_new=300)
dc1.remove_background()

# Plot after processing
axes[1].imshow(dc1.cube[10], cmap='viridis', aspect='auto')
axes[1].set_title("After Processing DC1")

# Save template for reuse
dc1.save_template("template.yml")

# show data
plt.tight_layout()
plt.show()

The figure below shows the effect of preprocessing. The left image represents a slice of the data cube before processing, while the right image shows the same slice after applying the recorded operations.

Visualizing the preprocessing steps on synthetic data.

Use a Template

Once a template has been created, it can be applied to new data sets to ensure consistent preprocessing. The example below demonstrates how to load and apply a saved template to another DataCube.

import wizard
from wizard._utils.example import generate_pattern_stack
import matplotlib.pyplot as plt

fig, axes = plt.subplots(1, 2, figsize=(10, 8))

# Apply template to new DataCube
# Create synthetic data and initialize DataCube
data2 = generate_pattern_stack(20, 300, 300, n_circles=10, n_rects=0, n_triangles=0, seed=43)
dc2 = wizard.DataCube(data2)

# Plot before processing
axes[0].imshow(dc2.cube[10], cmap='viridis', aspect='auto')
axes[0].set_title("Before Processing DC2")

dc2.execute_template("template.yml")

# Plot after processing
axes[1].imshow(dc2.cube[10], cmap='viridis', aspect='auto')
axes[1].set_title("After Processing DC2")

# Show final plot
plt.tight_layout()
plt.show()

The figure below shows the results of applying the template to a new data set.

Applying a saved processing template for consistent analysis.