QStudioMetrics
QStudioMetrics is a software to develop data mining and multivariate analysis studies.
Discoveries often arise from exploratory data analysis and come from a hypothesis-free manner alongside hypothesis-driven approaches. QStudioMetrics is a toolbox that allows a conversation between you, the data, and the hypotheses. It will enable you to explore and query the data and integrate your personal experience for new discoveries.
QStudioMetrics is written in C++ using the Qt5 framework and run under Linux, Windows and OSX.
QStudioMetrics compute:
- Principal Component Analysis (PCA)
- Consensus Principal Component Analysis (CPCA)
- Partial Least Squares (PLS)
- Multiple Linear Regression (MLR)
- Linear Discriminant Analysis (LDA)
- Clustering analysis: Hierarchical and K-Means
- Model validation: leave-one-out, bootstrap k-fold cross validation, y-scrambling
All the algorithms are part of libscientific (See https://github.com/gmrandazzo/libscientific)
Author: Giuseppe Marco Randazzo
Mantainer: Giuseppe Marco Randazzo, gmrandazzo@gmail.com
License
QStudioMetrics is distributed under LGPLv3 license. For more details please read the file "LICENSE" or go to "http://www.gnu.org/licenses/lgpl-3.0.html"
Install
Dependencies
- c++ compiler
- cmake >= 3.22
- Qt6
- libscientific >= 1.4.x
Compile from source
OSX
cmake .. -Wno-dev -DCMAKE_PREFIX_PATH=/usr/local/Cellar/qt/5.11.1/ -DLIBSCIENTIFIC_ROOT_DIR=/usr/local/Cellar/libscientific/HEAD-b4a8a4f/
make
Install binary release
Binary releases for Windows and OSX at https://github.com/gmrandazzo/QStudioMetrics/releases.
Documentation
https://qstudiometrics.readthedocs.io/en/latest/
Supported platforms
- Linux
- Windows
- Mac intel/silicon
Known bugs
- Hierarchical clustering: the problem belongs to the original libscientific library