Analysing scanning electron diffraction data using HyperSpy and pyXem

The Jupyter Notebooks in this deposit was originally used for the "ESTEEM3 Workshop: Electron diffraction for solving engineering problems" workshop held at the Norwegian University of Science and Technology (NTNU) in Trondheim 21.-23. June 2022. The aim of these notebooks is to give an introduction on how to work with very large scanning electron diffraction (SED) datasets acquired using a Transmission Electron Microscope (TEM), assuming no prior experience with Python, HyperSpy, pyXem or Jupyter Notebooks.

There are 5 notebooks included, which are intended to be done in order, starting with 01:

• 01_using_jupyter_notebooks: this gives an introduction on the Jupyter Notebook format itself, and how to use JupyterLab
• 02_working_with_diffraction_data: how to work with 4-dimensional scanning electron diffraction data, utilizing HyperSpy's lazy functionality to explore and process large datasets
• 03_centering_electron_beam: how to correct for the movement of the direct beam
• 04_advanced_data_processing_using_map: how to write custom functions for processing your data, by using HyperSpy's map functionality. Here by showing how different amorphous materials can be found in a SED dataset
• 05_dask_and_chunking: explores dataset chunking, which is an important concept when working with these large datasets

Prerequisites

To use these, you need to have JupyterLab, HyperSpy, and pyXem installed. The easiest way to do this, is to use the HyperSpy Bundle: https://hyperspy.org/hyperspy-doc/current/user_guide/install.html

These Jupyter Notebooks were made and tested using:

• hyperspy 1.7.1
• hyperspy-gui-traitsui 1.5.2
• hyperspy-gui-ipywidgets 1.5.0
• pyxem 0.14.2
• jupyterlab 3.4.2

However, they're likely to work with newer versions of these modules.

Datasets

There are three datasets from two different samples used in these notebooks. These are compressed in a 7zip-file, and needs to be extracted before they can be used with the notebooks. You might need install the open source 7zip software to extract this:

• Scanning Electron Diffraction from a Central Processing Unit (CPU), prepared for TEM using Focused Ion Beam (FIB) (cpu_sample_sped.zspy, cpu_sample_sped_small.hspy)
• Scanning Electron Diffraction from a permalloy film deposited on a SiN TEM grid, then irradiated by Ga-ions using a FIB (nanocrystalline_dscan.zspy)

The large files are (cpu_sample_sped.zspy, nanocrystalline_dscan.zspy) are stored in the `zspy` format http://hyperspy.org/hyperspy-doc/current/user_guide/io.html#zspy-hyperspy-s-zarr-specification, which is HyperSpy's specification of the Zarr format https://zarr.readthedocs.io/en/stable/.

Compared to the previous "standard" way of storing HyperSpy data, which was `hspy` (HDF5), `zspy` is much faster to both read and write.

Note

Not all the Jupyter Notebooks used in the "ESTEEM3 Workshop: Electron diffraction for solving engineering problems" is included in this deposit. This deposit includes the introductory ones presented by Magnus Nord.