# American Institute of Mathematical Sciences

March  2021, 3(1): 67-97. doi: 10.3934/fods.2021006

## HERMES: Persistent spectral graph software

 1 Department of Mathematics, Michigan State University, MI 48824, USA 2 Department of Computer Science and Engineering, Michigan State University, MI 48824, USA 3 Department of Mathematics, Department of Electrical and Computer Engineering, Department of Biochemistry and Molecular Biology, Michigan State University, MI 48824, USA

* Corresponding authors: Yiying Tong and Guo-Wei Wei

HERMES is available online here and via GitHub.

Received  December 2020 Revised  February 2021 Published  March 2021

Fund Project: This work was supported in part by NIH grant GM126189, NSF grants DMS-2052983, DMS-1761320, and IIS-1900473, NASA grant 80NSSC21M0023, Michigan Economic Development Corporation, George Mason University award PD45722, Bristol-Myers Squibb 65109, and Pfizer

Persistent homology (PH) is one of the most popular tools in topological data analysis (TDA), while graph theory has had a significant impact on data science. Our earlier work introduced the persistent spectral graph (PSG) theory as a unified multiscale paradigm to encompass TDA and geometric analysis. In PSG theory, families of persistent Laplacian matrices (PLMs) corresponding to various topological dimensions are constructed via a filtration to sample a given dataset at multiple scales. The harmonic spectra from the null spaces of PLMs offer the same topological invariants, namely persistent Betti numbers, at various dimensions as those provided by PH, while the non-harmonic spectra of PLMs give rise to additional geometric analysis of the shape of the data. In this work, we develop an open-source software package, called highly efficient robust multidimensional evolutionary spectra (HERMES), to enable broad applications of PSGs in science, engineering, and technology. To ensure the reliability and robustness of HERMES, we have validated the software with simple geometric shapes and complex datasets from three-dimensional (3D) protein structures. We found that the smallest non-zero eigenvalues are very sensitive to data abnormality.

Citation: Rui Wang, Rundong Zhao, Emily Ribando-Gros, Jiahui Chen, Yiying Tong, Guo-Wei Wei. HERMES: Persistent spectral graph software. Foundations of Data Science, 2021, 3 (1) : 67-97. doi: 10.3934/fods.2021006
##### References:

show all references

##### References:
Illustration of Voronoi diagram, Delaunay triangulation, and Non-Delaunay triangulation. Left chart: The Voronoi diagram and its dual Delaunay triangulation. The points set is $P$ = {A, B, C, D, E} and the Delaunay is defined as $\text{DT}(P)$. The blue lines tessellate the plane into Voronoi cells. The red circle are the circumcircles of triangles in $\text{DT}(P)$. Right chart: A Non-Delaunay triangulation. Vertices E and D are in the green circumcircles, implying the right chart is an example of Non-Delaunay triangulation
Illustration of 2D Delaunay triangulation, alpha shapes, and alpha complexes for a set of 6 points A, B, C, D, E, and F. Top left: The 2D Delaunay triangulation. Top right: The alpha shape and alpha complex at filtration value $\alpha = 0.2$. Bottom left: The alpha shape and alpha complex at filtration value $\alpha = 0.6$. Bottom right: The alpha shape and alpha complex at filtration value $\alpha = 1.0$. Here, we use dark blue color to fill the alpha shape
that are generated from Gudhi and DioDe">Figure 3.  The persistent barcode for a set of points as illustrated in Figure 2 that are generated from Gudhi and DioDe
The 3D structures of C$_{20}$ and C$_{60}$. (a) C$_{20}$ molecule. A total of 12 pentagon rings can be found in C$_{20}$. (b) C$_{60}$ molecule. 12 pentagon rings and 20 hexoagon rings form the structure of C$_{60}$
) at different filtration values $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $r$ (unit: Å), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{r, 0.05}$, $\mathcal{L}_1^{r, 0.05}$, and $\mathcal{L}_1^{r, 0.05}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{r, 0.05}$, $\mathcal{L}_1^{r, 0.05}$, and $\mathcal{L}_2^{r, 0.05}$ from top to bottom.">Figure 5.  Illustration of the harmonic spectra (for Rips complex) $\beta_0^{r, 0.05}$, $\beta_0^{r, 0.05}$, and $\beta_2^{r, 0.05}$ (green curves from top chart to bottom chart) and the smallest non-zero eigenvalue $\lambda_0^{r, 0.05}$, $\lambda_1^{r, 0.05}$, and $\lambda_2^{r, 0.05}$ (yellow curves from top chart to bottom chart) of C$_{20}$ molecule (the bottom left chart in Fig. 9) at different filtration values $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $r$ (unit: Å), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{r, 0.05}$, $\mathcal{L}_1^{r, 0.05}$, and $\mathcal{L}_1^{r, 0.05}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{r, 0.05}$, $\mathcal{L}_1^{r, 0.05}$, and $\mathcal{L}_2^{r, 0.05}$ from top to bottom.
) at different filtration value $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: Å), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{\alpha, 0.05}$, $\mathcal{L}_1^{\alpha, 0.05}$, and $\mathcal{L}_1^{\alpha, 0.05}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{\alpha, 0.05}$, $\mathcal{L}_1^{\alpha, 0.05}$, and $\mathcal{L}_2^{\alpha, 0.05}$ from top to bottom.">Figure 6.  Illustration of the harmonic spectra (for alpha complex) $\beta_0^{\alpha, 0.05}$, $\beta_0^{\alpha, 0.05}$, and $\beta_2^{\alpha, 0.05}$ (green curves from top chart to bottom chart) and the smallest non-zero eigenvalue $\lambda_0^{\alpha, 0.05}$, $\lambda_1^{\alpha, 0.05}$, and $\lambda_2^{\alpha, 0.05}$ (yellow curves from top chart to bottom chart) of C$_{20}$ molecule (the bottom left chart in Fig. 9) at different filtration value $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: Å), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{\alpha, 0.05}$, $\mathcal{L}_1^{\alpha, 0.05}$, and $\mathcal{L}_1^{\alpha, 0.05}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{\alpha, 0.05}$, $\mathcal{L}_1^{\alpha, 0.05}$, and $\mathcal{L}_2^{\alpha, 0.05}$ from top to bottom.
) at different filtration value $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{r, 0.05}$, $\mathcal{L}_1^{r, 0.05}$, and $\mathcal{L}_1^{r, 0.05}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{r, 0.05}$, $\mathcal{L}_1^{r, 0.05}$, and $\mathcal{L}_2^{r, 0.05}$ from top to bottom.">Figure 7.  Illustration of the harmonic spectra $\beta_0^{r, 0.05}$, $\beta_0^{r, 0.05}$, and $\beta_2^{r, 0.05}$ (blue curves from top chart to bottom chart) and the smallest non-zero eigenvalue $\lambda_0^{r, 0.05}$, $\lambda_1^{r, 0.05}$, and $\lambda_2^{r, 0.05}$ (red curves from top chart to bottom chart) of C$_{60}$ molecule (the bottom left chart in Fig. 9) at different filtration value $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{r, 0.05}$, $\mathcal{L}_1^{r, 0.05}$, and $\mathcal{L}_1^{r, 0.05}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{r, 0.05}$, $\mathcal{L}_1^{r, 0.05}$, and $\mathcal{L}_2^{r, 0.05}$ from top to bottom.
) at different filtration value $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{\alpha, 0.05}$, $\mathcal{L}_1^{\alpha, 0.05}$, and $\mathcal{L}_1^{\alpha, 0.05}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{\alpha, 0.05}$, $\mathcal{L}_1^{\alpha, 0.05}$, and $\mathcal{L}_2^{\alpha, 0.05}$ from top to bottom.">Figure 8.  Illustration of the harmonic spectra $\beta_0^{\alpha, 0.05}$, $\beta_0^{\alpha, 0.05}$, and $\beta_2^{\alpha, 0.05}$ (green curves from top chart to bottom chart) and the smallest non-zero eigenvalue $\lambda_0^{\alpha, 0.05}$, $\lambda_1^{\alpha, 0.05}$, and $\lambda_2^{\alpha, 0.05}$ (yellow curves from top chart to bottom chart) of C$_{60}$ molecule (the bottom left chart in Fig. 9) at different filtration value $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{\alpha, 0.05}$, $\mathcal{L}_1^{\alpha, 0.05}$, and $\mathcal{L}_1^{\alpha, 0.05}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{\alpha, 0.05}$, $\mathcal{L}_1^{\alpha, 0.05}$, and $\mathcal{L}_2^{\alpha, 0.05}$ from top to bottom.
The alpha carbon network plots of 15 proteins: PDB IDs 1CCR, 1NKO, 1O08, 1OPD, 1QTO, 1R7J, 1V70, 1W2L, 1WHI, 2CG7, 2FQ3, 2HQK, 2PKT, 2VIM, and 5CYT from left to right and top to bottom. The color represents the normalized diagonal element of the accumulated Laplacian at each alpha carbon atom
) at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable">Figure 10.  Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 5CYT (the bottom left chart in Fig. 9) at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
) at different filtration value $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{\alpha, 0.5}$, $\mathcal{L}_1^{\alpha, 0.5}$, and $\mathcal{L}_1^{\alpha, 0.5}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{\alpha, 0.5}$, $\mathcal{L}_1^{\alpha, 0.5}$, and $\mathcal{L}_2^{\alpha, 0.5}$ from top to bottom">Figure 11.  Illustration of the harmonic spectra $\beta_0^{\alpha, 0.5}$, $\beta_0^{\alpha, 0.5}$, and $\beta_2^{\alpha, 0.5}$ (green curves from top chart to bottom chart) and the smallest non-zero eigenvalue $\lambda_0^{\alpha, 0.5}$, $\lambda_1^{\alpha, 0.5}$, and $\lambda_2^{\alpha, 0.5}$ (yellow curves from top chart to bottom chart) of PDB ID 5CYT (the bottom left chart in Fig. 9) at different filtration value $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{\alpha, 0.5}$, $\mathcal{L}_1^{\alpha, 0.5}$, and $\mathcal{L}_1^{\alpha, 0.5}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{\alpha, 0.5}$, $\mathcal{L}_1^{\alpha, 0.5}$, and $\mathcal{L}_2^{\alpha, 0.5}$ from top to bottom
(a) The 3D secondary structure of PDB ID 1O08. The blue, purple, and orange colors represent helix, sheet, and random coils of PDB ID 1O08. The ball represents the alpha carbon of PDB ID 1O08. (b) Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 1O08 at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are calculated only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents for the number of zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents for the non-zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
) at different filtration value $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{r, 0}$, $\mathcal{L}_1^{r, 0}$, and $\mathcal{L}_1^{r, 0}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{r, 0}$, $\mathcal{L}_1^{r, 0}$, and $\mathcal{L}_2^{r, 0}$ from top to bottom.">Figure 13.  Illustration of the harmonic spectra $\beta_0^{r, 0}$, $\beta_0^{r, 0}$, and $\beta_2^{r, 0}$ (blue curves from top chart to bottom chart) and the smallest non-zero eigenvalue $\lambda_0^{r, 0}$, $\lambda_1^{r, 0}$, and $\lambda_2^{r, 0}$ (red curves from top chart to bottom chart) of C$_{60}$ molecule with one atom shifted (the bottom left chart in Fig. 9) at different filtration value $\alpha$ calculated from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axes represents the number of zero eigenvalues of $\mathcal{L}_0^{r, 0}$, $\mathcal{L}_1^{r, 0}$, and $\mathcal{L}_1^{r, 0}$ from top to bottom, and the right-$y$-axes represents the first non-zero eigenvalue of $\mathcal{L}_0^{r, 0}$, $\mathcal{L}_1^{r, 0}$, and $\mathcal{L}_2^{r, 0}$ from top to bottom.
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 1CCR at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 1NKO at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 1OPD at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 1QTO at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 1R7J at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 1V70 at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 1W2L at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 1WHI at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 2CG7 at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 2FQ3 at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 2HQK at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 2PKT at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
Illustration of the harmonic spectra $\beta_q^{\alpha, 0}$ (blue curve) and the smallest non-zero eigenvalue $\lambda_q^{\alpha, 0}$ (red curve) of PDB ID 2VIM at different filtration value $\alpha$ when $q = 0, 1, 2$. The $\beta_q^{\alpha, 0}$ are calculated from Gudhi, DioDe, and HERMES, and $\lambda_q^{\alpha, 0}$ are obtained only from HERMES. Here, the $x$-axis represents the radius filtration value $\alpha$ (unit: $Å$), the left-$y$-axis represents the number of zero eigenvalues of $\mathcal{L}_q^{\alpha, 0}$, and the right-$y$-axis represents the first non-zero eigenvalue of $\mathcal{L}_q^{\alpha, 0}$. Note that the harmonic spectra from three methods are indistinguishable
The matrix representation of $q$-boundary operator and its $q$th-order persistent Laplacian with corresponding dimension, rank, nullity, and spectra from alpha complex $K_{0.6} \to K_{0.6}$
The matrix representation of $q$-boundary operator and its $q$th-order persistent Laplacian with corresponding dimension, rank, nullity, and spectra from alpha complex $K_{0.2} \to K_{0.6}$
 [1] George Siopsis. Quantum topological data analysis with continuous variables. Foundations of Data Science, 2019, 1 (4) : 419-431. doi: 10.3934/fods.2019017 [2] Tyrus Berry, Timothy Sauer. Consistent manifold representation for topological data analysis. Foundations of Data Science, 2019, 1 (1) : 1-38. doi: 10.3934/fods.2019001 [3] Matthew O. Williams, Clarence W. Rowley, Ioannis G. Kevrekidis. A kernel-based method for data-driven koopman spectral analysis. Journal of Computational Dynamics, 2015, 2 (2) : 247-265. doi: 10.3934/jcd.2015005 [4] Massimiliano Guzzo, Giancarlo Benettin. A spectral formulation of the Nekhoroshev theorem and its relevance for numerical and experimental data analysis. Discrete & Continuous Dynamical Systems - B, 2001, 1 (1) : 1-28. doi: 10.3934/dcdsb.2001.1.1 [5] Antonio Rieser. A topological approach to spectral clustering. Foundations of Data Science, 2021, 3 (1) : 49-66. doi: 10.3934/fods.2021005 [6] Sarah Constantin, Robert S. Strichartz, Miles Wheeler. Analysis of the Laplacian and spectral operators on the Vicsek set. Communications on Pure & Applied Analysis, 2011, 10 (1) : 1-44. doi: 10.3934/cpaa.2011.10.1 [7] Keonhee Lee, Ngoc-Thach Nguyen, Yinong Yang. Topological stability and spectral decomposition for homeomorphisms on noncompact spaces. Discrete & Continuous Dynamical Systems, 2018, 38 (5) : 2487-2503. doi: 10.3934/dcds.2018103 [8] Eric Babson and Dmitry N. Kozlov. Topological obstructions to graph colorings. Electronic Research Announcements, 2003, 9: 61-68. [9] Mirela Domijan, Markus Kirkilionis. Graph theory and qualitative analysis of reaction networks. Networks & Heterogeneous Media, 2008, 3 (2) : 295-322. doi: 10.3934/nhm.2008.3.295 [10] Vu Hoang Linh, Volker Mehrmann. Spectral analysis for linear differential-algebraic equations. Conference Publications, 2011, 2011 (Special) : 991-1000. doi: 10.3934/proc.2011.2011.991 [11] Mark F. Demers, Hong-Kun Zhang. Spectral analysis of the transfer operator for the Lorentz gas. Journal of Modern Dynamics, 2011, 5 (4) : 665-709. doi: 10.3934/jmd.2011.5.665 [12] Rafael Tiedra De Aldecoa. Spectral analysis of time changes of horocycle flows. Journal of Modern Dynamics, 2012, 6 (2) : 275-285. doi: 10.3934/jmd.2012.6.275 [13] Pooja Bansal, Aparna Mehra. Integrated dynamic interval data envelopment analysis in the presence of integer and negative data. Journal of Industrial & Management Optimization, 2021  doi: 10.3934/jimo.2021023 [14] Zhouchen Lin. A review on low-rank models in data analysis. Big Data & Information Analytics, 2016, 1 (2&3) : 139-161. doi: 10.3934/bdia.2016001 [15] Pankaj Sharma, David Baglee, Jaime Campos, Erkki Jantunen. Big data collection and analysis for manufacturing organisations. Big Data & Information Analytics, 2017, 2 (2) : 127-139. doi: 10.3934/bdia.2017002 [16] Runqin Hao, Guanwen Zhang, Dong Li, Jie Zhang. Data modeling analysis on removal efficiency of hexavalent chromium. Mathematical Foundations of Computing, 2019, 2 (3) : 203-213. doi: 10.3934/mfc.2019014 [17] Habibe Zare Haghighi, Sajad Adeli, Farhad Hosseinzadeh Lotfi, Gholam Reza Jahanshahloo. Revenue congestion: An application of data envelopment analysis. Journal of Industrial & Management Optimization, 2016, 12 (4) : 1311-1322. doi: 10.3934/jimo.2016.12.1311 [18] Mahdi Mahdiloo, Abdollah Noorizadeh, Reza Farzipoor Saen. Developing a new data envelopment analysis model for customer value analysis. Journal of Industrial & Management Optimization, 2011, 7 (3) : 531-558. doi: 10.3934/jimo.2011.7.531 [19] M. Pellicer, J. Solà-Morales. Spectral analysis and limit behaviours in a spring-mass system. Communications on Pure & Applied Analysis, 2008, 7 (3) : 563-577. doi: 10.3934/cpaa.2008.7.563 [20] Aslihan Demirkaya, Panayotis G. Kevrekidis, Milena Stanislavova, Atanas Stefanov. Spectral stability analysis for standing waves of a perturbed Klein-Gordon equation. Conference Publications, 2015, 2015 (special) : 359-368. doi: 10.3934/proc.2015.0359

Impact Factor: