A GUI program for fast analysis of powder X-ray diffraction Images. It provides the capability of calibrating,
creating masks, having pattern overlays and showing phase lines.
It is known to run on Windows, Mac and Linux. For optimal usage on a Windows machine it should be run with 64 bit
python. When used with 32 bit Dioptas occasionally crashes because of limited memory allocation possibilities.
Installation
Executables
Executable versions for Windows, Mac OsX and Linux (all 64bit) can be downloaded from:
The executable versions are self-contained folders, which do not need any python installation. Under Windows and Linux
the program can be started by running the executable (e.g. Dioptas.exe or Dioptas). Under MacOS X the folder will
contain an app folder which can be double-clicked for starting. Please accept the security prompts.
Unfortunately, I am currently unable to create working App for MacOS X. Users of this convoluted operation system need
to install dioptas using “pip” and then start it from the commandline. (see below)
Code
In order to make changes to the source code yourself or always get the latest development versions you need to install
the required python packages on your machine.
The easiest way to install the all the dependencies for Dioptas is to use the pip package manager:
pip install dioptas
and then run Dioptas by typing:
dioptas
in the commandline.
Running the Program from source
The easiest way to create a working environment for Dioptas is to use the poetry package manager. A new environment is
created by running:
poetry install
This will create a new environment with all the required packages. The environment can be activated by running:
poetry shell
Afterward the program can be started by running:
dioptas
In order to run the program without activating the environment, the program can be started by running:
poetry run dioptas
In case you want to run the Dioptas from source without poetry, you need to install the required packages yourself.
The packages are listed in the file pyproject.toml. The program can then be started by running:
Dioptas
A GUI program for fast analysis of powder X-ray diffraction Images. It provides the capability of calibrating, creating masks, having pattern overlays and showing phase lines.
Maintainer
Clemens Prescher (clemens.prescher@gmail.com)
Requirements
It is known to run on Windows, Mac and Linux. For optimal usage on a Windows machine it should be run with 64 bit python. When used with 32 bit Dioptas occasionally crashes because of limited memory allocation possibilities.
Installation
Executables
Executable versions for Windows, Mac OsX and Linux (all 64bit) can be downloaded from:
https://github.com/Dioptas/Dioptas/releases
The executable versions are self-contained folders, which do not need any python installation. Under Windows and Linux the program can be started by running the executable (e.g. Dioptas.exe or Dioptas). Under MacOS X the folder will contain an app folder which can be double-clicked for starting. Please accept the security prompts.
Unfortunately, I am currently unable to create working App for MacOS X. Users of this convoluted operation system need to install dioptas using “pip” and then start it from the commandline. (see below)
Code
In order to make changes to the source code yourself or always get the latest development versions you need to install the required python packages on your machine.
The easiest way to install the all the dependencies for Dioptas is to use the pip package manager:
and then run Dioptas by typing:
in the commandline.
Running the Program from source
The easiest way to create a working environment for Dioptas is to use the poetry package manager. A new environment is created by running:
This will create a new environment with all the required packages. The environment can be activated by running:
Afterward the program can be started by running:
In order to run the program without activating the environment, the program can be started by running:
In case you want to run the Dioptas from source without poetry, you need to install the required packages yourself. The packages are listed in the file
pyproject.toml
. The program can then be started by running: