Introduction

Introduction#

The purpose of these tutorials is to provide a comprehensive collection of examples targeting a wide range of scientists and engineers interested in ptychography, from experimental users to algorithm and code developers.

Overview#

The table below gives an overview of all examples with associated data and a link for downloading the data.

Tutorial	Download Data	File Size	Reference
The Basics
The Parameter Tree	No Data Required	-	-
JSON/YAML Config Files	No Data Required	-	-
Input/Output Parameters	No Data Required	-	-
Scan Models	No Data Required	-	-
Reconstruction Engines	No Data Required	-	-
Projectional Engines	No Data Required	-	-
Stochastic Engines	No Data Required	-	-
Gradient-based Engines	No Data Required	-	-
Experimental X-ray Data
Loading HDF5 data	small_data.zip	126 MB	DOI
Working With Large Data	dls_i08_nanogold_spiral.zip	1.11 GB	DOI
Fixing Issues Related to Data (Loading)	small_data.zip	126 MB	DOI
Partial Coherence	dls_i08_nanogold_spiral.zip	1.11 GB	DOI
Position Refinement	dls_i14_test_structure.zip	3.68 GB	DOI
Missing Detector Frames	dls_i08_nanogold_raster.zip	993 MB	DOI
Testing different algorithms	dls_i13_butterfly.zip	286 MB	DOI
Using multiple GPUs	dls_i08_nanogold_spiral.zip	1.11 GB	DOI
Working with Data from SOLEIL - the SWING beamline	TBA	4.8 GB	TBA
Maximum Likelihood with the wavefield pre-conditioner	TBA	4.8 GB	TBA
Working with Data from NanoMAX	NanoMax example dataset 10_keV.zip	432 MB	DOI
Working with Data from SoftiMAX	ptycho_on_siemensstar_MAXIV_SoftiMAX.zip	565 MB	DOI
Working with Data from CoSAXS	ptycho_on_siemensstar_MAXIV_CoSAXS.zip	463 MB	DOI
Electron Ptychography Data
Working With Electron Data	dls_epsic_80kV_graphene.zip	270 MB	DOI
Chaining Multiple Engines	dls_epsic_80kV_graphene.zip	270 MB	DOI
Starting From a Previous Reconstruction	dls_epsic_80kV_graphene.zip	270 MB	DOI
Advanced Examples
Nearfield Ptychography	esrf_id16_AlNi_nearfield.zip	532 MB	TBA
Modifying an Engine	esrf_id16_AlNi_nearfield.zip	532 MB	TBA
Live visualisation	dls_i08_nanogold_spiral.zip	1.11 GB	DOI
Simulating Data	No Data Required	-	-
Modified Initial Probe	small_data.zip	126 MB	DOI
User Examples
Template	-	-	-
Super resolution	SiemensStar_ID16A_ESRF_farfieldptycho.zip	67 MB	DOI

Jupyter Lab/Notebooks#

Note

Most of the content of this tutorial has been created in the form of Jupyter Notebooks and the JupyterLab interface is the recommended way of working through these examples. JupyterLab is very easy to use, should you wish to learn more about the Jupyter project and the tools it provides, visit: https://jupyter.org.

As a minimal introduction, a few words about Jupyter Notebooks, which offer a convenient way to provide executable code (e.g. Python) alongside formatted text (using markdown). All notebooks used here are structured in the same way, with some informative content at the top and a full code example at the bottom, separated by a horizontal line:

In order to execute the code cell (or any other cell), simply select the cell and type SHIFT+ENTER or ALT+ENTER

Working with this tutorial within an HPC environment#

It might not always be possible to run Jupyter Notebooks with access to accelerated GPU hardware, for example in the case of only having terminal-based access to an HPC cluster. Below are a few pointers as to how best to run the PtyPy tutorials on an interactive HPC compute node, e.g. using Slurm.

Start an interactive session in the terminal, e.g. single GPU on slurm: srun --ntasks 1 --cpus-per-task 1 --gpus 1 --time 02:00:00 --pty bash
Create a new Python file and paste the full-length code block at the end of the Jupyter-based tutorial into it
Some of the settings in the PtyPy parameter tree are optimised for running it as a Jupyter Notebook, in particular the following should be modified when running interactively with Python

# Give your reconstruction a name
p.run = "test"

# Set verbose level to info
p.verbose_level = "info"

# Set io settings (no files saved)
p.io = u.Param()
p.io.home = "./"
p.io.rfile = "recons/%(run)s/%(run)s_%(engine)s_%(iterations)04d.ptyr"
p.io.autosave = u.Param(active=False)
p.io.interaction = u.Param(active=False)
p.io.autoplot = u.Param(active=False)

this will turn off all interactive plotting that is configured in all Jupyter Notebook tutorials, and it will save the results in a folder called recons in the folder defined as p.io.home.

How to visualise PtyPy output from a .ptyr file#

From a terminal with Graphics#

The simplest way to inspect the reconstruction results saved in the .ptyr file is to use the following command from a terminal with access to Graphics

ptypy.plot <PATH-TO-PTYR-FILE>

From a Jupyter Notebook or a Python script#

An alternative approach is to visualise the results in a Jupyter Notebook or from a Python script using the following code

import os
import ptypy.utils.plot_client as pc
ptyr_filename = "<PATH-TO-PTYR-FILE>"
fig = pc.figure_from_ptyr(ptyr_filename)