vincent-maillou commited on
Commit
5d1e260
·
1 Parent(s): f4dfce2

initial dataset commit

Browse files

Signed-off-by: vincent-maillou <vmaillou@iis.ee.ethz.ch>

.gitattributes CHANGED
@@ -57,3 +57,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
57
  # Video files - compressed
58
  *.mp4 filter=lfs diff=lfs merge=lfs -text
59
  *.webm filter=lfs diff=lfs merge=lfs -text
 
 
 
57
  # Video files - compressed
58
  *.mp4 filter=lfs diff=lfs merge=lfs -text
59
  *.webm filter=lfs diff=lfs merge=lfs -text
60
+ # Matrix market files
61
+ *.mtx filter=lfs diff=lfs merge=lfs -text
README.md CHANGED
@@ -1,3 +1,2 @@
1
- ---
2
- license: bsd-3-clause
3
- ---
 
1
+ # NEST: NEw Sparse maTrix dataset
2
+ The matrices are stored in the Matrix Market format as well as `scipy.sparse.npz` format.
 
device_modeling/README.md ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Device Modeling
2
+ This section of the dataset contains sparse matrices related to device modeling. More specifically, this contains matrices related to the modeling of semiconductor devices, including MOSFETs, ReRAM, and other devices.
3
+
4
+ ## kmc_
5
+ Domain:
6
+
7
+ ### Caracteristics
8
+ - Type:
9
+ - Size: , nnz
10
+
11
+ | | Initial | LU | Fill-in | Depth |
12
+ |----------|---------|----|---------|-------|
13
+ | | | | | |
14
+ | RCM | | | | |
15
+ | AMD | | | | |
16
+ | ND | | | | |
17
+
18
+ ### References
19
+ - [Accelerated Atomistic Kinetic Monte Carlo Simulations of Resistive Memory Arrays](https://ieeexplore.ieee.org/abstract/document/10793135)
20
+
21
+
22
+
23
+ ## qbit_
24
+ Domain:
25
+
26
+ ### Caracteristics
27
+ - Type:
28
+ - Size: , nnz
29
+
30
+ | | Initial | LU | Fill-in | Depth |
31
+ |----------|---------|----|---------|-------|
32
+ | | | | | |
33
+ | RCM | | | | |
34
+ | AMD | | | | |
35
+ | ND | | | | |
36
+
37
+ ### References
legacy_references/README.md ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Legacy References
2
+ This section of the dataset contains references sparse matrices that have been used as standard benchmarks for permutation and fill-in reduction algorithms as well as for evaluating the performance of sparse matrix factorization algorithms.
3
+
4
+ ## west0479
5
+ Domain: Chemical engineering plant models
6
+
7
+ ### Caracteristics
8
+ - Type: real unsymmetric
9
+ - Size: 479x479, 1910 nnz
10
+
11
+ | | Initial | LU | Fill-in | Depth |
12
+ |----------|---------|----|---------|-------|
13
+ | west0479 | 1910 | | | |
14
+ | RCM | | | | |
15
+ | AMD | | | | |
16
+ | ND | | | | |
17
+
18
+ ### References
19
+ - [State-of-The-Art Sparse Direct Solvers](https://arxiv.org/abs/1907.05309)
20
+
21
+
22
+ ## Freescale/circuit5M_dc
23
+ Domain: Circuit Simulation Problem
24
+
25
+ ### Caracteristics
26
+ - Type: real unsymmetric
27
+ - Size: 3523317x3523317, 19194193 nnz
28
+
29
+ | | Initial | LU | Fill-in | Depth |
30
+ |--------------|----------|----|---------|-------|
31
+ | circuit5M_dc | 19194193 | | | |
32
+ | RCM | | | | |
33
+ | AMD | | | | |
34
+ | ND | | | | |
35
+
36
+ ### References
37
+ - []()
legacy_references/circuit5m_dc/circuit5m_dc.mtx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:435d665495cdc1e7e6afb8e4d61cab905c479d6b0820d2a7f52d5bfb7b65d42e
3
+ size 621990406
legacy_references/circuit5m_dc/circuit5m_dc.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5f14a80342ca185a89448cf01f1ce1d52a6806bb11312887a9dc7584fd86714
3
+ size 31888094
legacy_references/west0479/west0479.mtx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a45b04df5fc8b27c6e87dba0fae80f734267d4c44deb5313578f5893a6a6122b
3
+ size 29246
legacy_references/west0479/west0479.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cfdb7f59e58475050d93516111620596ec0bee230a15df5e5c10eb458900ee2
3
+ size 13461
quantum_transport/README.md ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Quantum Transport
2
+ This section of the dataset contains sparse matrices related to quantum transport using the non-equilibrium Green's function method (NEGF).
3
+
4
+ ## cnt_cp2k
5
+ Domain: Quantum transport
6
+ Description: Hamiltonian (CP2K) of a carbon nano-tube made of 704 carbon atoms. Every carbon atom has 13 basis (9152 basis in total).
7
+
8
+ ### Caracteristics
9
+ - Type: complex symmetric
10
+ - Size: 9152x9152, 6154350 nnz
11
+
12
+ | | Initial | LU | Fill-in | Depth |
13
+ |----------|---------|----|---------|-------|
14
+ | cnt_cp2k | 6154350 | | | |
15
+ | RCM | | | | |
16
+ | AMD | | | | |
17
+ | ND | | | | |
18
+
19
+ ### References
20
+
21
+
22
+ ## cnt_w90
23
+ Domain: Quantum transport
24
+ Description: Hamiltonian (Wannier 90) of a carbon nano-tube.
25
+
26
+ ### Caracteristics
27
+ - Type: complex symmetric
28
+ - Size: 768x768, 71680 nnz
29
+
30
+ | | Initial | LU | Fill-in | Depth |
31
+ |----------|---------|----|---------|-------|
32
+ | cnt_cp2k | 71680 | | | |
33
+ | RCM | | | | |
34
+ | AMD | | | | |
35
+ | ND | | | | |
36
+
37
+ ### References
38
+
39
+
40
+ ## sinw_w90
41
+ Domain: Quantum transport
42
+ Description: Hamiltonian (Wannier 90) of a silicon nano-wire.
43
+
44
+ ### Caracteristics
45
+ - Type: complex symmetric
46
+ - Size: 7488x7488, 5310160 nnz
47
+
48
+ | | Initial | LU | Fill-in | Depth |
49
+ |----------|---------|----|---------|-------|
50
+ | sinw_w90 | 5310160 | | | |
51
+ | RCM | | | | |
52
+ | AMD | | | | |
53
+ | ND | | | | |
54
+
55
+ ### References
quantum_transport/cnt_cp2k/cnt_cp2k.mtx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e23a6569ea80456970ea0bdb6ec01741262f6ce990189a5d811d0176c60c643
3
+ size 193277075
quantum_transport/cnt_cp2k/cnt_cp2k.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c12515b163f89c64ec940dff7d994c9ec6c777563d2c586223da60f6755989c
3
+ size 48785865
quantum_transport/cnt_w90/cnt_w90.mtx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f55250c6a09de05d4d1524505dda1e37288bddd82961316c879973b68cc77b0c
3
+ size 1668885
quantum_transport/cnt_w90/cnt_w90.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a49650281d17c28276cb1220903edf6cf6c879d8b4ed5fe0cce0d615f5f01477
3
+ size 337563
quantum_transport/sinw_w90/sinw_w90.mtx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7151f429673d6c8b24d53f8fb719e402eddaec182bdbe14d0b0fb01c79cd78c
3
+ size 112328599
quantum_transport/sinw_w90/sinw_w90.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e9608d54b3ae97b6629a7bb8e1728db0bf455b8002a854efff34b9a2507cdc5
3
+ size 33100999
statistical_modeling/README.md ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Statistical Modeling
2
+ This section of the dataset contains sparse matrices related to statistical modeling. More specifically, this contain precision matrices as used within the integrated nested Laplace approximation (INLA) framework.
3
+
4
+ ## temp_prior_st1
5
+ Domain: Prior precision matrix from spatio-temporal statistical modeling applied to temperature prediction
6
+
7
+ ### Caracteristics
8
+ - Type: positive-definite symmetric
9
+ - Size: 1000506x1000506, 50870502 nnz
10
+
11
+ | | Initial | Cholesky | Fill-in | Depth |
12
+ |----------------|----------|----------|---------|-------|
13
+ | temp_prior_st1 | 50870502 | | | |
14
+ | RCM | | | | |
15
+ | AMD | | | | |
16
+ | ND | | | | |
17
+
18
+ ### References
19
+ - [Integrated Nested Laplace Approximations for Large-Scale Spatiotemporal Bayesian Modeling](https://epubs.siam.org/doi/full/10.1137/23M1561531)
20
+
21
+ ## temp_conditional_st1
22
+ Domain: Conditional precision matrix from spatio-temporal statistical modeling applied to temperature prediction
23
+
24
+ ### Caracteristics
25
+ - Type: positive-definite symmetric
26
+ - Size: 1000506x1000506, 62837532 nnz
27
+
28
+ | | Initial | Cholesky | Fill-in | Depth |
29
+ |----------------------|----------|----------|---------|-------|
30
+ | temp_conditional_st1 | 62837532 | | | |
31
+ | RCM | | | | |
32
+ | AMD | | | | |
33
+ | ND | | | | |
34
+
35
+ ### References
36
+ - [Integrated Nested Laplace Approximations for Large-Scale Spatiotemporal Bayesian Modeling](https://epubs.siam.org/doi/full/10.1137/23M1561531)
37
+
38
+
39
+
40
+ ## airpoll_prior_st3
41
+ Domain: Prior precision matrix from a coregional (trivariate) spatio-temporal statistical modeling applied to air pollution
42
+
43
+ ### Caracteristics
44
+ - Type: positive-definite symmetric
45
+ - Size: 8499x8499, 1205931 nnz
46
+
47
+ | | Initial | Cholesky | Fill-in | Depth |
48
+ |-------------------|---------|----------|---------|-------|
49
+ | airpoll_prior_st3 | 1205931 | | | |
50
+ | RCM | | | | |
51
+ | AMD | | | | |
52
+ | ND | | | | |
53
+
54
+ ### References
55
+ - []()
56
+ - [Multivariate disaggregation modeling of air pollutants: a case-study of PM2.5, PM10 and ozone prediction in Portugal and Italy](https://arxiv.org/abs/2503.12394)
57
+
58
+
59
+ ## airpoll_conditional_st3
60
+ Domain: Conditional precision matrix from a coregional (trivariate) spatio-temporal statistical modeling applied to air pollution
61
+
62
+ ### Caracteristics
63
+ - Type: positive-definite symmetric
64
+ - Size: 8499x8499, 1218651 nnz
65
+
66
+ | | Initial | Cholesky | Fill-in | Depth |
67
+ |-------------------------|---------|----------|---------|-------|
68
+ | airpoll_conditional_st3 | 1218651 | | | |
69
+ | RCM | | | | |
70
+ | AMD | | | | |
71
+ | ND | | | | |
72
+
73
+ ### References
74
+ - []()
75
+ - [Multivariate disaggregation modeling of air pollutants: a case-study of PM2.5, PM10 and ozone prediction in Portugal and Italy](https://arxiv.org/abs/2503.12394)
statistical_modeling/airpoll_conditional_st3/airpoll_conditional_st3.mtx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e6f36049f89c80adb3124452af96a3d3205347ac2e0e103d17a2331af7d5fc2
3
+ size 37475269
statistical_modeling/airpoll_conditional_st3/airpoll_conditional_st3.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc5a8da73869b4b2c29586cc86537329d8e1d7fb9287aea527781485803368fe
3
+ size 9266830
statistical_modeling/airpoll_prior_st3/airpoll_prior_st3.mtx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a89826dba4fa2c5d5b6012ac65ecfeb7c75692e65ed784f927d301b1e2176786
3
+ size 37106419
statistical_modeling/airpoll_prior_st3/airpoll_prior_st3.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:927af5993c8c77b7412aeb4fa6a02506904fa9d53ad59ce3144d857040c2e624
3
+ size 9189167
statistical_modeling/temp_conditional_st1/temp_conditional_st1.mtx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e40c1ad45cf901e6f19f89f5c3803251ce1fcccada921e413e0100b4f7a3349c
3
+ size 2183393090
statistical_modeling/temp_conditional_st1/temp_conditional_st1.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a7c6cb3798aea0a82869b6c1a3fa1924d1f9a50702475cd26ad6655290426d0
3
+ size 261009786
statistical_modeling/temp_prior_st1/temp_prior_st1.mtx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51fa08fce079591e883975f990d7f0d350d9e40c5096a7aac07435f4315887b8
3
+ size 1765512327
statistical_modeling/temp_prior_st1/temp_prior_st1.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:abd81801efd4f5ad5cbf62d21e8f95a07cb08b1b9ed0d039eac9b00056ef754f
3
+ size 143511377
utils/README.md ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Utils
2
+
3
+ ## matutils.py
4
+ This module contains utility functions for matrix handling and conversion from/to matrix market coo format and scipy npz format.
5
+
6
+ Example usage:
7
+ ```Python
8
+ # Load the west0479.mtx file, convert it to npz format and save it to the specified directory
9
+ python matutils.py mm2npz /home/vmaillou/Documents/Repositories/bigbird_dataset/legacy_references/west0479/west0479.mtx /home/vmaillou/Documents/Repositories/bigbird_dataset/ legacy_references/west0479/clear
10
+ ```
11
+
12
+ ```Python
13
+ # Load the west0479.npz file, convert it to Matrix Market format and save it to the specified directory
14
+ python matutils.py npz2mm /home/vmaillou/Documents/Repositories/bigbird_dataset/legacy_references/west0479/west0479.npz /home/vmaillou/Documents/Repositories/bigbird_dataset/legacy_references/west0479/
15
+ ```
utils/matutils.py ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os
2
+ import argparse
3
+ from pathlib import Path
4
+ import scipy as sp
5
+
6
+ def print_matinfos(matrix: sp.sparse.coo_matrix) -> None:
7
+ print(f" - Size of the matrix: {matrix.shape}")
8
+ print(f" - Number of non-zero elements: {matrix.nnz}")
9
+ print(f" - Data type of the matrix: {matrix.dtype}")
10
+
11
+
12
+ def load_matrixmarket(path_input: Path) -> sp.sparse.coo_matrix:
13
+ """
14
+ Load a Matrix Market file and return it as scipy.sparse.coo_matrix.
15
+ """
16
+ print(f" - Loaded Matrix Market file from: {path_input}")
17
+ matrix = sp.io.mmread(path_input)
18
+ print_matinfos(matrix)
19
+ return matrix
20
+
21
+ def save_matrixmarket(matrix: sp.sparse.coo_matrix, path_output: Path) -> None:
22
+ """
23
+ Save a scipy.sparse.coo_matrix to a Matrix Market file.
24
+ """
25
+ print(f" - Saving Matrix Market file to: {path_output}")
26
+ sp.io.mmwrite(
27
+ path_output,
28
+ matrix,
29
+ )
30
+ print(" - Successfully saved Matrix Market file.")
31
+
32
+
33
+ if __name__ == "__main__":
34
+ parser = argparse.ArgumentParser(description="Matrix Market utilities")
35
+ parser.add_argument("command", choices=["mm2npz", "npz2mm"], help="Command to execute")
36
+ parser.add_argument("input_file", help="Input file path")
37
+ parser.add_argument("output_file", help="Output file path")
38
+
39
+ args = parser.parse_args()
40
+
41
+ print("args:", args)
42
+
43
+ if args.command == "mm2npz":
44
+ print("Converting Matrix Market to NPZ format:")
45
+ matrix_coo = load_matrixmarket(args.input_file)
46
+ input_matrix_name = os.path.splitext(os.path.basename(args.input_file))[0]
47
+ output_matrix_name = Path.joinpath(
48
+ Path(args.output_file), f"{input_matrix_name}.npz"
49
+ )
50
+ print(f" - Saving to NPZ format: {output_matrix_name}")
51
+ sp.sparse.save_npz(output_matrix_name, matrix_coo)
52
+ elif args.command == "npz2mm":
53
+ print("Converting NPZ to Matrix Market format:")
54
+ matrix_coo = sp.sparse.load_npz(args.input_file).tocoo()
55
+ print(f" - Loaded NPZ file from: {args.input_file}")
56
+ print_matinfos(matrix_coo)
57
+ input_matrix_name = os.path.splitext(os.path.basename(args.input_file))[0]
58
+ output_matrix_name = Path.joinpath(
59
+ Path(args.output_file), f"{input_matrix_name}.mtx"
60
+ )
61
+ print(f" - Saving to Matrix Market format: {output_matrix_name}")
62
+ save_matrixmarket(matrix_coo, output_matrix_name)