diff --git a/joss.05420/10.21105.joss.05420.crossref.xml b/joss.05420/10.21105.joss.05420.crossref.xml new file mode 100644 index 0000000000..e0bd29d0fc --- /dev/null +++ b/joss.05420/10.21105.joss.05420.crossref.xml @@ -0,0 +1,316 @@ + + + + 20231025T124122-e97c05673a505ab57f757d9de77da11bb8a57c0e + 20231025124122 + + JOSS Admin + admin@theoj.org + + The Open Journal + + + + + Journal of Open Source Software + JOSS + 2475-9066 + + 10.21105/joss + https://joss.theoj.org + + + + + 10 + 2023 + + + 8 + + 90 + + + + libscientific: A Powerful C Library for Multivariate +Analysis + + + + Giuseppe Marco + Randazzo + https://orcid.org/0000-0003-1585-0019 + + + + 10 + 25 + 2023 + + + 5420 + + + 10.21105/joss.05420 + + + http://creativecommons.org/licenses/by/4.0/ + http://creativecommons.org/licenses/by/4.0/ + http://creativecommons.org/licenses/by/4.0/ + + + + Software archive + 10.5281/zenodo.8436823 + + + GitHub review issue + https://github.com/openjournals/joss-reviews/issues/5420 + + + + 10.21105/joss.05420 + https://joss.theoj.org/papers/10.21105/joss.05420 + + + https://joss.theoj.org/papers/10.21105/joss.05420.pdf + + + + + + Partial least-squares regression: A +tutorial + Geladi + Analytica Chimica Acta + 185 + 10.1016/0003-2670(86)80028-9 + 0003-2670 + 1986 + Geladi, P., & Kowalski, B. R. +(1986). Partial least-squares regression: A tutorial. Analytica Chimica +Acta, 185, 1–17. +https://doi.org/10.1016/0003-2670(86)80028-9 + + + Multi-way principal components-and +PLS-analysis + Wold + Journal of Chemometrics + 1 + 1 + 10.1002/cem.1180010107 + 0886-9383 + 1987 + Wold, S., Geladi, P., Esbensen, K., +& Öhman, J. (1987). Multi-way principal components-and PLS-analysis. +Journal of Chemometrics, 1(1), 41–56. +https://doi.org/10.1002/cem.1180010107 + + + Analysis of multiblock and hierarchical PCA +and PLS models + Westerhuis + Journal of Chemometrics + 5 + 12 + 10.1002/(SICI)1099-128X(199809/10)12:5<301::AID-CEM515>3.0.CO;2-S + 1998 + Westerhuis, J. A., Kourti, T., & +MacGregor, J. F. (1998). Analysis of multiblock and hierarchical PCA and +PLS models. Journal of Chemometrics, 12(5), 301–321. +https://doi.org/10.1002/(SICI)1099-128X(199809/10)12:5<301::AID-CEM515>3.0.CO;2-S + + + Multivariate analysis of quality : An +introduction + Martens + 10.1088/0957-0233/12/10/708 + 9780471974284 + 2001 + Martens, H., & Martens, M. +(2001). Multivariate analysis of quality : An introduction. Wiley. +https://doi.org/10.1088/0957-0233/12/10/708 + + + Statistical analysis with missing +data + Little + 10.1002/9781119013563 + 0471183865 + 1987 + Little, R. J. A., & Rubin, D. B. +(1987). Statistical analysis with missing data. Wiley. +https://doi.org/10.1002/9781119013563 + + + Robust L1 norm factorization in the presence +of outliers and missing data by alternative convex +programming + Ke + 1 + 10.1109/CVPR.2005.309 + 2005 + Ke, Q., & Kanade, T. (2005). +Robust L1 norm factorization in the presence of outliers and missing +data by alternative convex programming. 1, 739–746 vol. 1. +https://doi.org/10.1109/CVPR.2005.309 + + + Parameter based methods for compound +selection from chemical databases + Hudson + Quantitative Structure-Activity +Relationships + 4 + 15 + 10.1002/qsar.19960150402 + 1996 + Hudson, B. D., Hyde, R. M., Rahr, E., +Wood, J., & Osman, J. (1996). Parameter based methods for compound +selection from chemical databases. Quantitative Structure-Activity +Relationships, 15(4), 285–289. +https://doi.org/10.1002/qsar.19960150402 + + + Definitions of "dissimilarity" for +dissimilarity-based compound selection + Holliday + SLAS Discovery + 3 + 1 + 10.1177/108705719600100308 + 2472-5552 + 1996 + Holliday, J. D., & Willett, P. +(1996). Definitions of "dissimilarity" for dissimilarity-based compound +selection. SLAS Discovery, 1(3), 145–151. +https://doi.org/10.1177/108705719600100308 + + + Prediction of retention time in +reversed-phase liquid chromatography as a tool for steroid +identification + Randazzo + Analytica Chimica Acta + 916 + 10.1016/j.aca.2016.02.014 + 0003-2670 + 2016 + Randazzo, G. M., Tonoli, D., Hambye, +S., Guillarme, D., Jeanneret, F., Nurisso, A., Goracci, L., Boccard, J., +& Rudaz, S. (2016). Prediction of retention time in reversed-phase +liquid chromatography as a tool for steroid identification. Analytica +Chimica Acta, 916, 8–16. +https://doi.org/10.1016/j.aca.2016.02.014 + + + Indirect quantitative structure-retention +relationship for steroid identification: A chemometric challenge at +“chimiométrie 2016” + Randazzo + Chemometrics and Intelligent Laboratory +Systems + 160 + 10.1016/j.chemolab.2016.11.010 + 0169-7439 + 2017 + Randazzo, G. M., Vigneau, E., +Courcoux, P., Harrouet, C., Lijour, Y., Dardenne, P., Boccard, J., & +Rudaz, S. (2017). Indirect quantitative structure-retention relationship +for steroid identification: A chemometric challenge at “chimiométrie +2016.” Chemometrics and Intelligent Laboratory Systems, 160, 52–58. +https://doi.org/10.1016/j.chemolab.2016.11.010 + + + Enhanced metabolite annotation via dynamic +retention time prediction: Steroidogenesis alterations as a case +study + Randazzo + Journal of Chromatography B + 1071 + 10.1016/j.jchromb.2017.04.032 + 1570-0232 + 2017 + Randazzo, G. M., Tonoli, D., +Strajhar, P., Xenarios, I., Odermatt, A., Boccard, J., & Rudaz, S. +(2017). Enhanced metabolite annotation via dynamic retention time +prediction: Steroidogenesis alterations as a case study. Journal of +Chromatography B, 1071, 11–18. +https://doi.org/10.1016/j.jchromb.2017.04.032 + + + Steroid identification via deep learning +retention time predictions and two-dimensional gas chromatography-high +resolution mass spectrometry + Randazzo + Journal of Chromatography A + 1612 + 10.1016/j.chroma.2019.460661 + 0021-9673 + 2020 + Randazzo, G. M., Bileck, A., Danani, +A., Vogt, B., & Groessl, M. (2020). Steroid identification via deep +learning retention time predictions and two-dimensional gas +chromatography-high resolution mass spectrometry. Journal of +Chromatography A, 1612, 460661. +https://doi.org/10.1016/j.chroma.2019.460661 + + + Dual therapeutic targeting of intra-articular +inflammation and intracellular bacteria enhances chondroprotection in +septic arthritis + Kwon + Science Advances + 26 + 7 + 10.1126/sciadv.abf2665 + 2021 + Kwon, H.-K., Lee, I., Yu, K. E., +Cahill, S. V., Alder, K. D., Lee, S., Dussik, C. M., Back, J., Choi, J., +Song, L., Kyriakides, T. R., & Lee, F. Y. (2021). Dual therapeutic +targeting of intra-articular inflammation and intracellular bacteria +enhances chondroprotection in septic arthritis. Science Advances, 7(26), +eabf2665. https://doi.org/10.1126/sciadv.abf2665 + + + Treating ‘septic’ with enhanced antibiotics +and ‘arthritis’ by mitigation of excessive inflammation + Kwon + Frontiers in Cellular and Infection +Microbiology + 12 + 10.3389/fcimb.2022.897291 + 2235-2988 + 2022 + Kwon, H.-K., Dussik, C. M., Kim, +S.-H., Kyriakides, T. R., Oh, I., & Lee, F. Y. (2022). Treating +“septic” with enhanced antibiotics and “arthritis” by mitigation of +excessive inflammation. Frontiers in Cellular and Infection +Microbiology, 12. +https://doi.org/10.3389/fcimb.2022.897291 + + + A theoretical foundation for the PLS +algorithm + Lorber + Journal of Chemometrics + 1 + 1 + 10.1002/cem.1180010105 + 1987 + Lorber, A., Wangen, L. E., & +Kowalski, B. R. (1987). A theoretical foundation for the PLS algorithm. +Journal of Chemometrics, 1(1), 19–31. +https://doi.org/10.1002/cem.1180010105 + + + + + + diff --git a/joss.05420/10.21105.joss.05420.jats b/joss.05420/10.21105.joss.05420.jats new file mode 100644 index 0000000000..23e4b8f2ff --- /dev/null +++ b/joss.05420/10.21105.joss.05420.jats @@ -0,0 +1,684 @@ + + +
+ + + + +Journal of Open Source Software +JOSS + +2475-9066 + +Open Journals + + + +5420 +10.21105/joss.05420 + +libscientific: A Powerful C Library for Multivariate +Analysis + + + +https://orcid.org/0000-0003-1585-0019 + +Randazzo +Giuseppe Marco + + + + + +Independent researcher + + + + +17 +3 +2023 + +8 +90 +5420 + +Authors of papers retain copyright and release the +work under a Creative Commons Attribution 4.0 International License (CC +BY 4.0) +2022 +The article authors + +Authors of papers retain copyright and release the work under +a Creative Commons Attribution 4.0 International License (CC BY +4.0) + + + +C +Python +chemometrics +multivariate analysis + + + + + + Summary +

Multivariate analysis is a powerful technique that allows + researchers to analyze and interpret data with multiple variables. In + today’s data-driven world, multivariate analysis has become essential + for the exploration of complex data sets. libscientific is a powerful + library written in C that provides a comprehensive set of multivariate + analysis tools based on the NIPALS algorithm. The library includes + several multivariate analysis algorithms, such as principal component + analysis (PCA), partial least squares regression (PLS), consensus + principal component analysis (CPCA), multiblock principal component + analysis, and multiblock partial least squares (UPLS). libscientific + also includes several other tools to analyze data, such as cluster + analysis using KMeans Hierarchical clustering and other methods to run + linear algebra calculations. The library also provides a Python + foreign function to be used inside Python scripts.

+ + + Statement of need +

The library is designed to be easy to use and can be integrated + into any C or C++ project. Additionally, libscientific comes with a + foreign function Python bindings, making it accessible within Python + scripts and easier to perform data analysis tasks. One of the main + advantages of libscientific is its performance and scalability. This + means that large data sets can be analyzed quickly and efficiently, + making it an ideal choice for applications where speed is critical. + The library depends only on lapack for SVD and eigenvalues + decomposition and can be easily integrated into embedded systems. The + current library version is 1.6.0, and here is a list of the current + library features:

+ + +

Principal Component Analysis (PCA)

+ + +

Consensus Principal Component Analysis (CPCA)

+ + +

Partial Least Squares (PLS)

+ + +

Multiple Linear Regression (MLR)

+ + +

Unfold Principal Component Analysis (UPCA)

+ + +

Unfold Partial Least Squares (UPLS)

+ + +

Fisher Linear Discriminant Analysis (LDA)

+ + +

Kmeans++ Clustering

+ + +

Hierarchical Clustering

+ + +

Sample selection algorithms: Most Descriptive Compound (MDC), + Most Dissimilar Compound (MaxDis)

+ + +

Statistical measures: R2, MSE, MAE, RMSE, Sensitivity, PPV

+ + +

Yates Analysis

+ + +

Receiver Operating Characteristic curve anaysis (ROC)

+ + +

Precision-Recal curve analysis

+ + +

Matrix-matrix Euclidean, Manhattan, Cosine and Mahalanobis + distances

+ + +

Numerical integration

+ + +

Natural cubic spline interpolation and prediction

+ + +

Linear algebra (Eigenvector/value and SVD operated by Lapack + library)

+ + +

Ordinary Least Squares solver

+ + +

Linear equation Solver

+ + +

Nelder-Mead Simplex Optimization

+ + +

Cross validation methods: Bootstrap k-fold, Leave-One-Out, + Y-Scrambling

+ + +

libscientific was designed to analyze any kind of multivariate + tabular data and to be applied in any scientific field.

+ + + State of The field +

The primary objective of libscientific is to offer a library + capable of performing multivariate analysis by implementing the NIPALS + algorithm. This choice stems from the limitations of prevailing + popular libraries like scikit-learn, which need help handling missing + and noisy data effectively + (Ke + & Kanade, 2005; + Little + & Rubin, 1987). Notably, the NIPALS algorithm is a robust + solution to these challenges, addressing issues related to data + incompleteness without necessitating prior data imputation. + Furthermore, the NIPALS algorithm conducts iterative computations of + components and latent variables, leading to more efficient use of + memory resources than alternative methods found in scikit-learn. + Notably, it demonstrates superior computational efficiency, especially + when researchers seek to analyze only a select few of the foremost + principal components or latent variables. In summary, libscientific + aims to provide a sophisticated solution for multivariate analysis, + leveraging the NIPALS algorithm’s strengths to surmount issues related + to missing and noisy data, optimize component calculations, and + enhance computational efficiency in scenarios where the analysis + focuses on a limited number of critical principal components or latent + variables.

+ + + Multivariate analysis algorithms specs +

Principal component analysis (PCA) is one of the most commonly used + methods for multivariate analysis. PCA is an unsupervised method that + compresses data into low-dimensional representations that capture the + dominant variation in the data. libscientific provides a robust + implementation of PCA using the NIPALS algorithm described in Geladi + & Kowalski + (1986). + libscientific implementation can handle data sets with many variables, + few instances, and missing values.

+

Partial least squares (PLS) is another commonly used method for + multivariate analysis. PLS is a supervised method that captures the + dominant covariation between the data matrix and the target/response. + libscientific provides one version of PLS described by Geladi & + Kowalski + (1986). + This implementation works with single-task and multi-task regression + problems.

+

In addition to PCA and PLS, libscientific provides implementations + of Consensus PCA (CPCA) to analyze time series and multi-block data, + algorithm described by Westerhuis et al. + (1998), + and other multi-block methods such as Unfold Principal Component + Analysis (UPCA) and Unfold Partial Least Squares (UPLS) both + implementation from Wold et al. + (1987).

+

All multivariate algorithms admit missing values since the core + linear algebra functions are coded to skip missing values, according + to Martens & Martens + (2001), + p. 381.

+ + + Other algorithms +

The library also provides compound selection algorithms such as + Most Descriptive Compounds + (Hudson + et al., 1996) or Most Dissimilar Compound + (Holliday + & Willett, 1996) selections, allowing one to analyze scores + plots or original data matrices and select samples based on the + object/sample diversity. Moreover, multi-thread cross-validation + methodologies such as “Bootstrap k-fold” Leave-One-Out (LOO), and + Y-Scrambling tests are implemented to facilitate the scientist in + testing model prediction abilities.

+ + + Algorithm stability +

Since we are dealing with numerical analysis, unit tests are + crucial to ensure correctness, stability, and reproducibility. + libscientific tests range from simple matrix-vector multiplication to + the correctness of complex algorithms using ad-hoc torture toy + examples. The code coverage is reported to be more than 75%, + indicating that a larger portion of the code has been verified to work + as expected, reducing the likelihood of undiscovered bugs. This is + important since libscientific is a set of implementations of + algorithms that involves complex mathematical calculation, and + correctness and accuracy are crucial in minimizing the risk of + numerical errors in scientific, engineering, and data analysis + applications.

+ + + Speed and Memory Comparison +

Several simulations of every algorithm in libscientific with data + of different sizes (input size) against CPU speed were performed to + address the algorithm’s performance.

+

Looking at the plots for PCA, CPCA, and PLS in Figure 1, we observe + a linear trend, which suggests that the algorithm’s time complexity + also increases linearly. However, this does not tell the computational + complexity of the algorithms. Indeed, since the NIPALS algorithm is + similar to the power method for determining the eigenvectors and + eigenvalues of a matrix + (Lorber + et al., 1987), we can assume that the computational complexity + could be O(n²) or O(n³). Moreover it is important to point out that + the calculation speed is mainly influenced by the number of samples, + the number of iterations required for convergence, and the number of + principal components/latent variables to calculate.

+

Instead, MLR shows a polynomial correlation as expected from the + OLS algorithm, which uses a matrix direct inverse approach with a + computational complexity of O(n³).

+

With this analysis, we confirm that as the input size (often termed + “problem size”) increases by a constant factor, the execution time + also increases proportionally (linear algorithms). Linear algorithms + have notable characteristics:

+ + +

Most of the time, ‘Linear Time Complexity (O(n))’: Execution + time grows linearly with input size.

+ + +

Constant Work per Input Element: Each input element is + processed continuously in linear algorithms.

+ + +

Stable Performance Impact: Doubling input size roughly doubles + execution time, facilitating performance estimation.

+ + +

Optimal Scaling: Linear-time solutions efficiently handle + larger inputs.

+ + + + + + + + + + + + + + + + + + + + + + + + + + +
PCACPCA
+ PCA + + CPCA +
PLSMLR
+ PLS + + MLR +
+ +

Figure 1: Speed performances of 4 different algorithms: Principal + Component Analysis (PCA), Partial Least-Squares (PLS), Consensus + Principal Component Analysis (CPCA), and Multiple Linear Regression + (MLR). Simulations reveal linear trends for PCA, CPCA, and PLS, + hinting at probable linear time complexity. However, it is worth + mentioning that the NIPALS algorithm, influenced by sample size, + iterations, latent variables, and similar to the power method for + estimating eigenvectors/eigenvalues, may report O(n²) or O(n³) + complexity. Meanwhile, MLR exhibits a polynomial correlation typical + of the OLS matrix approach with O(n³) complexity.

+ + + Usage +

For the usage in C or either Python we invite reading the official + documentation located at the following link: + http://gmrandazzo.github.io/libscientific/

+ + + Conclusions +

libscientific offers a potent suite of multivariate analysis tools + that greatly enhance the ability of researchers and analysts to + extract valuable insights from diverse tabular data. With its robust + C-based implementation and seamless Python bindings, the library + balances high performance and user-friendliness, making it an optimal + solution for swiftly executing data-driven applications.

+

Incorporating libscientific into analytical workflows may empower + professionals to leverage various multivariate techniques to crack + complex relationships and patterns within datasets. By offering tools + for data reduction, predictive modeling, quality control, and more, as + already demonstrated in previous works in -omics science and + predictive modeling + (Kwon + et al., 2021, + 2022; + Randazzo + et al., 2016, + 2020; + Randazzo, + Vigneau, et al., 2017; + Randazzo, + Tonoli, et al., 2017), the library can be an indispensable + asset for tackling intricate challenges across various + disciplines.

+ + + Acknowledgements +

libscientific was born as an open-source project from the + Ph.D. thesis of the author Giuseppe Marco Randazzo. The author + acknowledges the support from the University of Perugia, the valuable + code review made by the people from Freaknet Medialab, and the bug + reports from the whole open-source community using this library.

+ + + + + + + + GeladiPaul + KowalskiBruce R. + + Partial least-squares regression: A tutorial + Analytica Chimica Acta + 1986 + 185 + 0003-2670 + https://www.sciencedirect.com/science/article/pii/0003267086800289 + 10.1016/0003-2670(86)80028-9 + 1 + 17 + + + + + + WoldSvante + GeladiPaul + EsbensenKim + ÖhmanJerker + + Multi-way principal components-and PLS-analysis + Journal of Chemometrics + John Wiley & Sons, Ltd + 19870101 + 1 + 1 + 0886-9383 + 10.1002/cem.1180010107 + 41 + 56 + + + + + + WesterhuisJohan A. + KourtiTheodora + MacGregorJohn F. + + Analysis of multiblock and hierarchical PCA and PLS models + Journal of Chemometrics + 1998 + 12 + 5 + https://analyticalsciencejournals.onlinelibrary.wiley.com/doi/abs/10.1002/%28SICI%291099-128X%28199809/10%2912%3A5%3C301%3A%3AAID-CEM515%3E3.0.CO%3B2-S + 10.1002/(SICI)1099-128X(199809/10)12:5<301::AID-CEM515>3.0.CO;2-S + 301 + 321 + + + + + + MartensHarald + MartensM. + + Multivariate analysis of quality : An introduction + Wiley + Chichester + 2001 + 9780471974284 + https://www.wiley.com/en-us/Multivariate+Analysis+of+Quality%3A+An+Introduction-p-9780471974284 + 10.1088/0957-0233/12/10/708 + + + + + + LittleRoderick J. A. + RubinDonald B. + + Statistical analysis with missing data + Wiley + 1987 + 0471183865 + https://onlinelibrary.wiley.com/doi/book/10.1002/9781119013563 + 10.1002/9781119013563 + + + + + + KeQifa + KanadeT. + + Robust L1 norm factorization in the presence of outliers and missing data by alternative convex programming + 2005 + 1 + + 10.1109/CVPR.2005.309 + 739 + 746 vol. 1 + + + + + + HudsonBrian D. + HydeRichard M. + RahrElizabeth + WoodJohn + OsmanJulian + + Parameter based methods for compound selection from chemical databases + Quantitative Structure-Activity Relationships + 1996 + 15 + 4 + https://onlinelibrary.wiley.com/doi/abs/10.1002/qsar.19960150402 + 10.1002/qsar.19960150402 + 285 + 289 + + + + + + HollidayJohn D. + WillettPeter + + Definitions of "dissimilarity" for dissimilarity-based compound selection + SLAS Discovery + Elsevier + 19960401 + 1 + 3 + 2472-5552 + 10.1177/108705719600100308 + 145 + 151 + + + + + + RandazzoGiuseppe Marco + TonoliDavid + HambyeStephanie + GuillarmeDavy + JeanneretFabienne + NurissoAlessandra + GoracciLaura + BoccardJulien + RudazSerge + + Prediction of retention time in reversed-phase liquid chromatography as a tool for steroid identification + Analytica Chimica Acta + 2016 + 916 + 0003-2670 + https://www.sciencedirect.com/science/article/pii/S000326701630215X + 10.1016/j.aca.2016.02.014 + 8 + 16 + + + + + + RandazzoGiuseppe Marco + VigneauEvelyne + CourcouxPhilippe + HarrouetCorentin + LijourYves + DardennePierre + BoccardJulien + RudazSerge + + Indirect quantitative structure-retention relationship for steroid identification: A chemometric challenge at “chimiométrie 2016” + Chemometrics and Intelligent Laboratory Systems + 2017 + 160 + 0169-7439 + https://www.sciencedirect.com/science/article/pii/S0169743916302143 + 10.1016/j.chemolab.2016.11.010 + 52 + 58 + + + + + + RandazzoGiuseppe Marco + TonoliDavid + StrajharPetra + XenariosIoannis + OdermattAlex + BoccardJulien + RudazSerge + + Enhanced metabolite annotation via dynamic retention time prediction: Steroidogenesis alterations as a case study + Journal of Chromatography B + 2017 + 1071 + 1570-0232 + https://www.sciencedirect.com/science/article/pii/S157002321730716X + 10.1016/j.jchromb.2017.04.032 + 11 + 18 + + + + + + RandazzoGiuseppe Marco + BileckAndrea + DananiAndrea + VogtBruno + GroesslMichael + + Steroid identification via deep learning retention time predictions and two-dimensional gas chromatography-high resolution mass spectrometry + Journal of Chromatography A + 2020 + 1612 + 0021-9673 + https://www.sciencedirect.com/science/article/pii/S0021967319310830 + 10.1016/j.chroma.2019.460661 + 460661 + + + + + + + KwonHyuk-Kwon + LeeInkyu + YuKristin E. + CahillSean V. + AlderKareme D. + LeeSaelim + DussikChristopher M. + BackJungHo + ChoiJeongjoon + SongLee + KyriakidesThemis R. + LeeFrancis Y. + + Dual therapeutic targeting of intra-articular inflammation and intracellular bacteria enhances chondroprotection in septic arthritis + Science Advances + 2021 + 7 + 26 + https://www.science.org/doi/abs/10.1126/sciadv.abf2665 + 10.1126/sciadv.abf2665 + eabf2665 + + + + + + + KwonHyuk-Kwon + DussikChristopher M. + KimSang-Hun + KyriakidesThemis R. + OhIrvin + LeeFrancis Y. + + Treating ‘septic’ with enhanced antibiotics and ‘arthritis’ by mitigation of excessive inflammation + Frontiers in Cellular and Infection Microbiology + 2022 + 12 + 2235-2988 + https://www.frontiersin.org/articles/10.3389/fcimb.2022.897291 + 10.3389/fcimb.2022.897291 + + + + + + LorberAvraham + WangenLawrence E. + KowalskiBruce R. + + A theoretical foundation for the PLS algorithm + Journal of Chemometrics + 1987 + 1 + 1 + https://analyticalsciencejournals.onlinelibrary.wiley.com/doi/abs/10.1002/cem.1180010105 + 10.1002/cem.1180010105 + 19 + 31 + + + + +
diff --git a/joss.05420/10.21105.joss.05420.pdf b/joss.05420/10.21105.joss.05420.pdf new file mode 100644 index 0000000000..bda7eb3940 Binary files /dev/null and b/joss.05420/10.21105.joss.05420.pdf differ diff --git a/joss.05420/media/performance/cpca_input_vs_cputime.png b/joss.05420/media/performance/cpca_input_vs_cputime.png new file mode 100644 index 0000000000..ff78dae34c Binary files /dev/null and b/joss.05420/media/performance/cpca_input_vs_cputime.png differ diff --git a/joss.05420/media/performance/mlr_input_vs_cputime.png b/joss.05420/media/performance/mlr_input_vs_cputime.png new file mode 100644 index 0000000000..318285f3bf Binary files /dev/null and b/joss.05420/media/performance/mlr_input_vs_cputime.png differ diff --git a/joss.05420/media/performance/pca_input_vs_cputime.png b/joss.05420/media/performance/pca_input_vs_cputime.png new file mode 100644 index 0000000000..f06e6a9bec Binary files /dev/null and b/joss.05420/media/performance/pca_input_vs_cputime.png differ diff --git a/joss.05420/media/performance/pls_input_vs_cputime.png b/joss.05420/media/performance/pls_input_vs_cputime.png new file mode 100644 index 0000000000..22437f8362 Binary files /dev/null and b/joss.05420/media/performance/pls_input_vs_cputime.png differ