SkillAgentSearch skills...

EvenTem

4D STEM directly on raw ASI CheeTah T3, Advacam TPX3 or QD Merlin data

Install / Use

/learn @EMAT-Jo/EvenTem
About this skill

Quality Score

0/100

Supported Platforms

Universal

README

EvenTem

EvenTem started as a fork of riCOM, but has grown to a complete tool for fast 4D STEM reconstruction, with a focus on live update capabilities and optimization for event-driven pipelines. EvenTem is fully written in C++, but has a python API through pybind11, so that it can be simply installed and used like any python module.

API documentation can be found here.

Supported detectors are

as well as data in the following formats

  • Frame-based numpy files (dtype np.uint8 & np.uint16, shape [RX,RY,K,K] with K either 64,128,256,512)
  • Event-based .electron (custom dtype [RX,RY,KX,XY,scan number], all uint16)

Installation

Pre-compiled binaries should be available for Windows, Linux and Apple silicon supporting python 3.9 to 3.12. For these, installation requires to clone the repository using

git clone https://github.com/EMAT-Jo/EvenTem.git

and then to navigate to the EvenTem directory and use

pip install .

Testing the installation can be done as

import EvenTem

GUI Use

see EvenTem GUI repo

Scripting Use

An example reconstruction requires as little as

from EvenTem import vSTEM
vs = vSTEM(nx=N,ny=N,repetitions=repetitions,filename=filename)
vs.DetectorSize = DetectorSize
vs.InnerRadia = [vs.DetectorSize//2]
vs.OuterRadia = [vs.DetectorSize]
vs.Run()
vs.PlotImage()

#results are numpy accessible
result = vs.image #np.array

For complete example notebooks, check the examples directory.

View on GitHub
GitHub Stars8
CategoryDevelopment
Updated1mo ago
Forks1

Languages

C++

Security Score

85/100

Audited on Feb 6, 2026

No findings