ASTRA - Stellar Astrophysics Data Manager

ASTRA is a modern Qt6 application for managing and analyzing stellar astrophysics data. It provides comprehensive tools for handling large catalogs of stars, spectroscopic data, photometry, radial velocities, and SEDs.

Quick Start

If you want to immediately get going using ASTRA, the by far quickest way is grabbing the AppImage file from the latest release you can see on the right side of the GitHub page. Once its downloaded on your Linux system, navigate to its location and launch it using

chmod +x  ./astra-0.3.0-x86_64.AppImage
./astra-0.3.0-x86_64.AppImage --appimage-extract-and-run

(Change the version number to match your downloaded file)

Installation

ASTRA can be installed by downloading the latest AppImage release (recommended for most users, see above), or built from source.

Runtime Dependencies (AppImage users)

The AppImage bundles Qt and most libraries, but a few system packages must be installed separately:

Debian / Ubuntu (22.04+):

sudo apt install libfuse2 libopengl0 libegl1 libxkbcommon0 libgl1 \
                 libdbus-1-3 libfontconfig1 \
                 python3 python3-numpy gnuplot

Arch Linux:

sudo pacman -S fuse2 libglvnd libxkbcommon dbus fontconfig \
               python python-numpy gnuplot

Then:

chmod +x astra-0.2.2-x86_64.AppImage
./astra-0.2.2-x86_64.AppImage

---

Building from Source

ASTRA is a CMake project requiring Qt ≥ 6.10, a Fortran compiler, and several scientific libraries.

Automated installer (Ubuntu / Debian / Arch)

install-linux.sh does everything in this section for you: it detects your distribution, installs the packages, downloads Qt via aqtinstall when your distro is too old, fetches the git submodules, builds the optional lcurve fitting binaries, and finally builds and installs ASTRA.

git clone https://github.com/Fabmat1/ASTRA.git
cd ASTRA
./install-linux.sh

Useful options: --prefix ~/.local (per-user install, no root needed), --build-only, --no-lcurve, --skip-deps, --no-cuda, -j N, -v. Run ./install-linux.sh --help for the full list. If you would rather do it by hand, follow the steps below.

1. Install Build Dependencies

Ubuntu 22.04+ / Debian 12+:

sudo apt install build-essential cmake git wget pkg-config gfortran \
                 qt6-base-dev qt6-tools-dev libqt6opengl6-dev \
                 libqt6svg6-dev qt6-declarative-dev \
                 libgl1-mesa-dev libxkbcommon-dev libvulkan-dev \
                 libeigen3-dev libopenblas-dev liblapack-dev \
                 libboost-all-dev nlohmann-json3-dev \
                 libfftw3-dev libtbb-dev libcxxopts-dev \
                 libccfits-dev libcfitsio-dev \
                 python3-dev python3-numpy gnuplot

Installing Qt 6.10+ on Ubuntu 22.04 / 24.04:

pip install aqtinstall
aqt install-qt linux desktop 6.11.1 linux_gcc_64 -O ~/Qt

# Make CMake aware of it
export CMAKE_PREFIX_PATH=$HOME/Qt/6.11.1/gcc_64:$CMAKE_PREFIX_PATH
export PATH=$HOME/Qt/6.11.1/gcc_64/bin:$PATH

Add the two export lines to your ~/.bashrc if you want them persistent.

Arch Linux:

sudo pacman -S base-devel cmake git wget pkgconf gcc-fortran \
               qt6-base qt6-tools qt6-svg qt6-declarative \
               vulkan-headers vulkan-icd-loader \
               eigen openblas lapack boost nlohmann-json \
               fftw onetbb cxxopts \
               ccfits cfitsio \
               python python-numpy gnuplot

2. Header-only Dependencies

unordered_dense is not packaged on most distros - install it manually:

sudo mkdir -p /usr/local/include/ankerl
sudo wget -O /usr/local/include/ankerl/unordered_dense.h \
  https://raw.githubusercontent.com/martinus/unordered_dense/v4.4.0/include/ankerl/unordered_dense.h
sudo ln -sf /usr/local/include/ankerl/unordered_dense.h /usr/local/include/unordered_dense.h

3. Clone and Build

git clone --recurse-submodules https://github.com/Fabmat1/ASTRA.git
cd ASTRA

cmake -S . -B build \
  -DCMAKE_BUILD_TYPE=Release \
  -DGAEL_ENABLE_CUDA=OFF   # set ON if you have CUDA installed

cmake --build build -j$(nproc)

The compiled binary will be at build/ASTRA. Run it directly:

./build/ASTRA

4. (Optional) System-wide Install

To install ASTRA system-wide - which also registers desktop files so it appears in your application menu:

sudo cmake --install build --prefix /usr/local

After install, launch it from your application menu or run ASTRA from any terminal.

License

MIT License

AI Declaration

ASTRA was made using AI assistance, but without agentic or autonomous coding. All code was carefully reviewed and pushed manually.

Further instructions will be added to this README in the future - hang tight!