Pyzfn

Pyzfn is a micromagnetic post-processing library designed for analyzing and visualizing data generated by amumax, a fork of mumax3. It provides a convenient interface for working with simulation data stored in Zarr format, offering optimized functions for data manipulation, spectral analysis, and visualization.

Features

• Easy Data Access: Load and navigate large simulation datasets efficiently.
• Spectral Analysis: Calculate dispersion relations and extract spin wave modes.
• Visualization Tools: Generate snapshots and interactive spectral plots.
• Integration with Zarr: Leverage Zarr's capabilities for handling big data.

Installation

Install Pyzfn using pip:

pip install pyzfn

Usage

Pyzfn builds upon the functionalities provided by Zarr. Familiarity with Zarr is recommended for advanced usage.

Importing Pyzfn

from pyzfn import Pyzfn

Loading a Simulation

# Replace 'path/to/job.zarr' with your simulation data path
job = Pyzfn('path/to/job.zarr')

Exploring the Data Structure

Pretty Print the Data Tree

# Interactive tree (requires IPython)
job.pp

Simple Data Tree Print

# Non-interactive print
job.p

Accessing Simulation Metadata

Simulation metadata is stored in the .attrs attribute:

# Access metadata as a dictionary
dx = job.attrs['dx']

# Or as a class attribute (if no dataset shares the same name)
dx = job.dx

Retrieving Simulation Data

Access datasets saved during the simulation:

# Access magnetization data
# For example, get the x-component of magnetization for time steps 10 <= t < 20
# Dimensions are (time, z, y, x, component)
mag = job.m[10:20, :, :, :, 0]  # mag is a NumPy array

Lazy Loading

To handle large datasets without loading them into memory immediately:

# Get a reference to the dataset
mag_dset = job.m

# Load data when needed
arr = mag_dset[:]

Dataset Information

Display detailed information about a dataset:

mag_dset.info

Utilizing Help Functions

Use Python's built-in help and dir functions for assistance:

help(mag_dset)

Post-Processing Functions

Pyzfn provides optimized functions for common post-processing tasks.

Calculating Dispersion Relations

# Calculate the dispersion relation for the magnetization dataset
job.calc_disp(dset_in_str='m', dset_out_str='m')

Calculating Spin Wave Modes

# Extract spin wave modes from the data
job.calc_modes(dset_in_str='m', dset_out_str='m')

Data Visualization

Snapshot Visualization

Generate a snapshot of the magnetization at a specific time step:

# Visualize the magnetization at time index t=50 and layer z=0
job.snapshot(t=50, z=0)

Interactive Spectral Analysis

Launch an interactive FFT spectrum plot:

# Interactive spectrum analysis
job.ispec(dset='m', c=0)

Advanced Usage

Custom Slicing

Optimize memory usage and performance with custom slicing:

# Define custom slices for processing
slices = (
    slice(None),  # Time dimension
    slice(None),  # z dimension
    slice(None),  # y dimension
    slice(None),  # x dimension
    slice(None)   # Component dimension
)

# Calculate modes with custom slices
job.calc_modes(dset_in_str='m', dset_out_str='m', slices=slices)

Saving and Loading OVF Files

Pyzfn includes utilities for working with OOMMF's OVF file format:

from pyzfn.utils import save_ovf, load_ovf

# Save data to an OVF file
save_ovf('output.ovf', mag)

# Load data from an OVF file
data = load_ovf('output.ovf')

Trimming Modes

Trim modes based on peak detection to reduce data size:

# Trim modes with specified parameters
job.trim_modes(dset_in_str='m', dset_out_str='trimmed_m', peak_xcut_min=0)

Contributing

Contributions are welcome! If you find a bug or have a feature request, please open an issue or submit a pull request on the GitHub repository.

Acknowledgments

• amumax for simulation data.
• Zarr for efficient data handling.