ShiftExperiment CLI Builder

Weights Mode

ImageNet weights is the default behavior if no subcommand is provided. Choose a subcommand mode below.

mode

Dataset Configuration

--source_dir (required)

Root directory for the source distribution.

--target_dir (required)

Root directory for the target distribution.

--source_list_path

List file for source images.

--target_list_path (required)

List file for target images.

Data Perturbations

Artificially shift the target data.

--gaussian_sigma

Noise std dev. (0.0=off)

--rotation_angle

Angle in degrees.

--shear_angle

Shear angle.

--width_shift_frac

Fraction of width.

--height_shift_frac

Fraction of height.

--zoom_factor

Scale (1.0 = none).

--crop_image

--horizontal_flip

--vertical_flip

Sampling & Test Controls

Controls sample size, run count, and statistical-test settings.

--sample_size

Images per run.

--num_runs

Number of calibration and test runs.

--block_idx

Index for block sampling.

--batch_size

Dataloader batch size.

--image_size

Input image size (pixels).

--latent_dim

Autoencoder latent dimensions.

--alpha

Significance level (0–1).

--seed_base

Base seed for runs.

--permutation_test_iterations

0 disables permutation p-values

Significance Preview

Percentile: 95.0th

Do not assume a Gaussian distribution. Gaussian shown for illustrative purposes only. The red line marks the (1 - alpha) * 100 percentile threshold.

Logging Output

Where to write JSON experiment logs (directory must exist).

--file_location

Directory path; must already exist.

--file_name

Output JSON filename.

Required: --source_dir, --target_dir, --target_list_path. If custom weights, also --model_weights_path.