https://github.com/hdarjus/sparvaride-matlab
Matlab Implementation of Algorithms for Sparse Factor Analyses
https://github.com/hdarjus/sparvaride-matlab
econometrics factor-analysis latent-factors matlab parameter-identification
Last synced: about 1 month ago
JSON representation
Matlab Implementation of Algorithms for Sparse Factor Analyses
- Host: GitHub
- URL: https://github.com/hdarjus/sparvaride-matlab
- Owner: hdarjus
- License: gpl-3.0
- Created: 2023-03-02T06:50:34.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2025-12-13T15:54:33.000Z (6 months ago)
- Last Synced: 2025-12-27T01:59:24.854Z (6 months ago)
- Topics: econometrics, factor-analysis, latent-factors, matlab, parameter-identification
- Language: MATLAB
- Homepage:
- Size: 31.3 KB
- Stars: 0
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Software package `sparvaride-matlab`
Matlab implementation of the main algorithm of the R package [`sparvaride`](hdarjus.github.io/sparvaride/) plus two utilitites for the generalized lower triangular (GLT) framework in our [paper published in Econometrics](https://doi.org/10.3390/econometrics11040026) by Sylvia Frühwirth-Schnatter, Darjus Hosszejni, and Hedibert Freitas Lopes.
The R package `sparvaride` implements the variance identification algorithm for sparse
factor analysis described in the paper "Cover It Up\! Bipartite Graphs
Uncover Identifiability in Sparse Factor Analysis" by Darjus Hosszejni
and Sylvia Frühwirth-Schnatter, published in Journal of Multivariate Analysis
(2026). https://doi.org/10.1016/j.jmva.2025.105536
## Functionality
There are three functions and two further Matlab files in this repository.
- `to_glt` rotates the input into GLT representation
- a Matlab-default numeric tolerance is applied
- this function is for *continuous* factor loadings matrices
- the transpose of the factor loadings matrix is the input
- further info in Section 3.3 of our [paper published in Econometrics](https://doi.org/10.3390/econometrics11040026)
- `is_identified_indicator` verifies the counting rule for the *sparsity pattern* of a factor loadings matrix
- there is no numeric error involved in this algorithm
- the sparsity pattern of the transpose of the factor loadings matrix is the input
- further info in our [paper](https://doi.org/10.1016/j.jmva.2025.105536), which is about what this algorithm is and why it works
- `is_identified` is the sequence of `to_glt` and `is_identified_indicator` on the sparsity pattern of the resulting GLT structure
## How to use
You need to have Matlab on your computer first; I tested it with Matlab R2021b.
Then, you have to download the `.m` files of this repository onto your computer and put them into your Matlab's working directory.
See `tests.m` in this repository for code examples.
Execute `run_tests.m` to run the unit tests.
## Citation
Please cite our papers if you use the code.
### Relevant paper for citing the `is_identified` function:
Darjus Hosszejni and Sylvia Frühwirth-Schnatter (2026). "Cover It Up! Bipartite Graphs Uncover Identifiability in Sparse Factor Analysis." Journal of Multivariate Analysis, 211, 105536. https://doi.org/10.1016/j.jmva.2025.105536
A BibTeX entry for LaTeX users is
@article{,
title = {Cover It Up! {Bipartite} Graphs Uncover Identifiability in Sparse Factor Analysis},
author = {Darjus Hosszejni and Sylvia Fr\"uhwirth-Schnatter},
journal = {Journal of Multivariate Analysis},
volume = {211},
pages = {105536},
year = {2026},
issn = {0047-259X},
doi = {10.1016/j.jmva.2025.105536},
url = {https://doi.org/10.1016/j.jmva.2025.105536}
}
### Relevant paper for citing the `to_glt` function:
Sylvia Frühwirth-Schnatter, Darjus Hosszejni, and Hedibert Freitas Lopes (2023). "When It Counts—Econometric Identification of the Basic Factor Model Based on GLT Structures." Econometrics 11(4), 26. https://doi.org/10.3390/econometrics11040026
A BibTex entry for LaTeX users is
@article{,
title = {When It Counts—{Econometric} Identification of the Basic Factor Model Based on {GLT} Structures},
author = {Sylvia Fr\"uhwirth-Schnatter and Darjus Hosszejni and Hedibert Freitas Lopes},
journal = {Econometrics},
volume = {11},
number = {4},
pages = {26},
year = {2023},
doi = {10.3390/econometrics11040026},
url = {https://doi.org/10.3390/econometrics11040026}
}
## Issues
If you do not like these references or you have other questions, open a ticket here in the [Issue tracker](https://github.com/hdarjus/sparvaride-matlab/issues).