ndslice

Quick interactive visualization for N-dimensional NumPy arrays

A python package for browsing slices, applying FFTs, and inspecting data.

Quickly checking multi-dimensional data usually means writing the same matplotlib boilerplate over and over. This tool lets you just call ndslice(data) and interactively explore what you've got.

Usage

from ndslice import ndslice
import numpy as np

# Create some data
x = np.linspace(-5, 5, 100)
y = np.linspace(-5, 5, 100)
z = np.linspace(-5, 5, 50)
X, Y, Z = np.meshgrid(x, y, z, indexing='ij')
mag = np.exp(-(X**2 + Y**2 + Z**2) / 10)
pha = np.pi/4 * (X + Y + Z)
complex_data = mag * np.exp(1j * pha)

ndslice(complex_data, title='3D Complex Gaussian')

Showcase

Features

Data slicing and dimension selection should be intuitive: click the two dimensions you want to show and slice using the spinboxes.

Centered FFT - Click dimension labels to apply centered 1D FFT transforms. Useful for checking k-space data in MRI reconstructions or analyzing frequency content. FFT

Line plot - See 1D slices through your data. Shift+scroll for Y zoom, Ctrl+scroll for X zoom:

Line plot

Video export Right-clicking a dimension button to export a video or PNG frames along that dimension. The video export functionality is optional, and can be installed with

pip install ndslice[video_export]

Export

Scaling

Log scaling is often good for k-space visualization. Symmetric log scaling is an extension of the log scale which supports negative values.

Colormap Change colormap:

Ctrl+1: Gray
Ctrl+2: Viridis
Ctrl+3: Plasma
Ctrl+4: Cyclic rainbow, hides phase wraps
Ctrl+5: Cividis
Ctrl+6: Cubehelix
Ctrl+7: Cool
Ctrl+8: Warm

Axis flipping Click arrow icons (⬇️/⬆️ and ⬅️/➡️) next to dimension labels to flip axes. Default orientation is image-style (origin lower-left). Flip the primary axis for matrix-style (origin upper-left).

Non-blocking windows

By default, windows open in separate processes, allowing multiple simultaneous views:

ndslice(data1)
ndslice(data2) # Both windows appear

Use block=True to wait for the window to close before continuing:

ndslice(data1, block=True)  # Script pauses here
ndslice(data2)  # Shown after first closes

Command Line

ndslice data.npy # Numpy file
ndslice --help   # Show all options

File support ndslice has CLI support and can conveniently display: | Format | File suffix | Requirement | |---|---:|---| | NumPy | .npy, .npz | NumPy | | MATLAB | .mat | scipy | | HDF5 | .h5, .hdf5 | h5py | | BART | .cfl + .hdr | — | | Philips REC | .REC + .xml | — | | NIfTI | .nii, .nii.gz | nibabel | | DICOM (pixel array) | .dcm | pydicom |

HDF5 files can be compound complex dtype, or real/imag fields.

If there are multiple datasets in the file, a selection GUI appears which highlights arrays supported by ndslice (essentially numeric). Double click to open.

Selector

Installation

From PyPI

pip install ndslice 
pip install ndslice[video_export] # Include video export dependencies (imageio, Pillow, imageio-ffmpeg)
pip install ndslice[all] # includes all optional dependencies (video export and file formats)

From source

git clone https://github.com/henricryden/ndslice.git
cd ndslice

# Use directly without installing
python -m ndslice data.npy

pip install -e .

Requirements

Python >= 3.8
NumPy >= 1.20.0
PyQtGraph >= 0.12.0
PyQt5 >= 5.15.0
h5py >= 3.0.0 (for HDF5 file support in CLI)

License

MIT License - see LICENSE file for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Acknowledgments

Built with PyQtGraph for high-performance visualization.

Henric Rydén

Karolinska University Hospital

Stockholm, Sweden

Ndslice

Install / Use

README