J/A+A/649/A53       Mean galaxy spectra of the 86 classes  (Fraix-Burnet+, 2021)

Unsupervised classification of SDSS galaxy spectra. Fraix-Burnet D., Bouveyron C., Moultaka J. <Astron. Astrophys. 649, A53 (2021)> =2021A&A...649A..53F 2021A&A...649A..53F (SIMBAD/NED BibCode)
ADC_Keywords: Galaxies, spectra ; Spectral types Keywords: methods: data analysis - methods: statistical - galaxies: statistics - galaxies: general - techniques: spectroscopic Abstract: Defining templates of galaxy spectra is useful to quickly characterise new obervations and organise data bases from surveys. These templates are usually built from a pre-defined classification based on other criteria. We present an unsupervised classification of 702248 spectra of galaxies and quasars with redshifts smaller than 0.25 that were retrieved from the Sloan Digital Sky Survey (SDSS) database, release 7. The spectra were first corrected for the redshift, then wavelet-filtered to reduce the noise, and finally binned to obtain about 1437 wavelengths per spectrum. Fisher-EM, an unsupervised clustering discriminative latent mixture model algorithm, was applied on these corrected spectra, considering the full set as well as several subsets of 100000 and 300000 spectra. The optimum number of classes given by a penalised likelihood criterion is 86 classes, the 37 most populated ones gathering 99% of the sample. These classes are established from a subset of 302214 spectra. Using several cross-validation techniques we find that this classification is in agreement with the results obtained on the other subsets with an average misclassification error of about 15%. The large number of very small classes tends to increase this error rate. In this paper, we make a first quick comparison of our classes with the templates of Kennicutt (1992), Dobos et al (2012), Wang et al (2018). This is the first time that an automatic, objective and robust unsupervised classification is established on such a large amount of spectra of galaxies. The mean spectra of the classes can be used as templates for a large majority of galaxies in our Universe. Description: The spectra used in this study were retrieved from the SDSS DR7 data base. They were corrected for redshift, denoised and rebinned as explained in the paper. They were NOT corrected for extinction. The full set of processed spectra is available on request. In the table classk86.dat, we provide the index specObjId of SDSS DR7. The mean spectra of the 86 classes is provided here in electronic form, as well as the class for each of the 302248 spectra used for this classification. File Summary: -------------------------------------------------------------------------------- FileName Lrecl Records Explanations -------------------------------------------------------------------------------- ReadMe 80 . This file meansk86.dat 1590 1437 Mean spectra of the 86 classes classk86.dat 30 302248 Classes K86 of 302248 spectra tablec1.dat 29 86 *Correspondence between our classes (Fisher-EM) and those of other atlases -------------------------------------------------------------------------------- Note on tablec1.dat: Kennicutt 1992ApJS...79..255K 1992ApJS...79..255K; Dobos et al. 2012MNRAS.420.1217D 2012MNRAS.420.1217D; Wang et al. 2018MNRAS.474.1873W 2018MNRAS.474.1873W -------------------------------------------------------------------------------- See also: http://skyserver.sdss.org/dr7 : SDSS DR7 Home Page Byte-by-byte Description of file: meansk86.dat -------------------------------------------------------------------------------- Bytes Format Units Label Explanations -------------------------------------------------------------------------------- 1- 16 F16.11 0.1nm lambda Wavelength (Angstroem) 18- 34 F17.15 --- A8 Mean spectra of class A8 36- 52 F17.15 --- A16 Mean spectra of class A16 54- 70 F17.15 --- A22 Mean spectra of class A22 72- 88 F17.15 --- A35 Mean spectra of class A35 90- 106 F17.15 --- A21 Mean spectra of class A21 108- 124 F17.15 --- A30 Mean spectra of class A30 126- 142 F17.15 --- A23 Mean spectra of class A23 144- 162 F19.16 --- A25 Mean spectra of class A25 164- 180 F17.15 --- A38 Mean spectra of class A38 182- 200 F19.16 --- A4 Mean spectra of class A4 202- 218 F17.15 --- A29 Mean spectra of class A29 220- 236 F17.15 --- A7 Mean spectra of class A7 238- 254 F17.15 --- A43 Mean spectra of class A43 256- 272 F17.15 --- A5 Mean spectra of class A5 274- 290 F17.15 --- A18 Mean spectra of class A18 292- 308 F17.15 --- A11 Mean spectra of class A11 310- 326 F17.15 --- A3 Mean spectra of class A3 328- 344 F17.15 --- A36 Mean spectra of class A36 346- 362 F17.15 --- A50 Mean spectra of class A50 364- 380 F17.15 --- A37 Mean spectra of class A37 382- 398 F17.15 --- A28 Mean spectra of class A28 400- 416 F17.15 --- A6 Mean spectra of class A6 418- 434 F17.15 --- A15 Mean spectra of class A15 436- 452 F17.15 --- A39 Mean spectra of class A39 454- 470 F17.15 --- A2 Mean spectra of class A2 472- 488 F17.15 --- A20 Mean spectra of class A20 490- 506 F17.15 --- A45 Mean spectra of class A45 508- 526 F19.17 --- A46 Mean spectra of class A46 528- 544 F17.15 --- A26 Mean spectra of class A26 546- 562 F17.15 --- A47 Mean spectra of class A47 564- 580 F17.15 --- A32 Mean spectra of class A32 582- 598 F17.15 --- A10 Mean spectra of class A10 600- 616 F17.15 --- A17 Mean spectra of class A17 618- 635 F18.15 --- A48 Mean spectra of class A48 637- 653 F17.15 --- A13 Mean spectra of class A13 655- 672 F18.16 --- A44 Mean spectra of class A44 674- 690 F17.15 --- A9 Mean spectra of class A9 692- 708 F17.15 --- A49 Mean spectra of class A49 710- 727 F18.15 --- A40 Mean spectra of class A40 729- 745 F17.15 --- A34 Mean spectra of class A34 747- 763 F17.15 --- A19 Mean spectra of class A19 765- 781 F17.15 --- A31 Mean spectra of class A31 783- 799 F17.15 --- A33 Mean spectra of class A33 801- 817 F17.15 --- A41 Mean spectra of class A41 819- 835 F17.15 --- A27 Mean spectra of class A27 837- 853 F17.15 --- A24 Mean spectra of class A24 855- 871 F17.15 --- A12 Mean spectra of class A12 873- 889 F17.15 --- A42 Mean spectra of class A42 891- 907 F17.15 --- A1 Mean spectra of class A1 909- 925 F17.15 --- A14 Mean spectra of class A14 927- 944 F18.15 --- B12 Mean spectra of class B12 946- 963 F18.16 --- B20 Mean spectra of class B20 965- 981 F17.15 --- B7 Mean spectra of class B7 983- 999 F17.15 --- B19 Mean spectra of class B19 1001-1017 F17.15 --- B14 Mean spectra of class B14 1019-1036 F18.15 --- B6 Mean spectra of class B6 1038-1055 F18.15 --- B24 Mean spectra of class B24 1057-1073 F17.15 --- B5 Mean spectra of class B5 1075-1093 F19.17 --- B21 Mean spectra of class B21 1095-1111 F17.15 --- B1 Mean spectra of class B1 1113-1129 F17.15 --- B16 Mean spectra of class B16 1131-1147 F17.15 --- B2 Mean spectra of class B2 1149-1165 F17.15 --- B13 Mean spectra of class B13 1167-1183 F17.15 --- B17 Mean spectra of class B17 1185-1201 F17.15 --- B8 Mean spectra of class B8 1203-1222 F20.17 --- B23 Mean spectra of class B23 1224-1240 F17.15 --- B15 Mean spectra of class B15 1242-1258 F17.15 --- B3 Mean spectra of class B3 1260-1276 F17.15 --- B4 Mean spectra of class B4 1278-1295 F18.16 --- B25 Mean spectra of class B25 1297-1313 F17.15 --- B10 Mean spectra of class B10 1315-1331 F17.15 --- B9 Mean spectra of class B9 1333-1349 F17.15 --- B22 Mean spectra of class B22 1351-1367 F17.15 --- B18 Mean spectra of class B18 1369-1385 F17.15 --- B11 Mean spectra of class B11 1387-1403 F17.15 --- C9 Mean spectra of class C9 1405-1422 F18.15 --- C7 Mean spectra of class C7 1424-1441 F18.15 --- C10 Mean spectra of class C10 1443-1460 F18.15 --- C1 Mean spectra of class C1 1462-1479 F18.15 --- C3 Mean spectra of class C3 1481-1498 F18.15 --- C6 Mean spectra of class C6 1500-1516 F17.15 --- C5 Mean spectra of class C5 1518-1534 F17.15 --- C8 Mean spectra of class C8 1536-1553 F18.15 --- C2 Mean spectra of class C2 1555-1571 F17.15 --- C4 Mean spectra of class C4 1573-1590 F18.15 --- D Mean spectra of class D -------------------------------------------------------------------------------- Byte-by-byte Description of file: classk86.dat -------------------------------------------------------------------------------- Bytes Format Units Label Explanations -------------------------------------------------------------------------------- 1- 18 I18 --- specObjID The SDSS index of the spectrum 20- 30 A11 --- FEMK86class The class of each spectrum (1) -------------------------------------------------------------------------------- Note (1): Class "Disregarded" indicates the 34 weird spectra that were not used in the classification. -------------------------------------------------------------------------------- Byte-by-byte Description of file: tablec1.dat -------------------------------------------------------------------------------- Bytes Format Units Label Explanations -------------------------------------------------------------------------------- 1- 3 A3 --- Class Fisher-EM class (this paper) (1) 5- 9 A5 --- K92 Kennicutt (1992ApJS...79..255K 1992ApJS...79..255K, Cat. VII/141) class 11- 16 A6 --- D12 Dobos et al. (2012MNRAS.420.1217D 2012MNRAS.420.1217D) class (2) 18- 29 A12 --- W18 Wang et al. (2018MNRAS.474.1873W 2018MNRAS.474.1873W, Cat. J/MNRAS/474/1873) class -------------------------------------------------------------------------------- Note (1): Our classes are ordered according to their number of spectra. Note (2): Dobos et al. (2012MNRAS.420.1217D 2012MNRAS.420.1217D) define the nomenclature as: p = passive l = LINER s = Seyfert h = Hα hh = AGN+HII t = all BG = all blue galaxies GG = all green galaxies RG = all red galaxies -------------------------------------------------------------------------------- Acknowledgements: Didier Fraix-Burnet, didier.fraix-burnet(at)univ-grenoble-alpes.fr
(End) D. Fraix-Burnet [IPAG/CNRS/UGA, France], P. Vannier [CDS] 19-Feb-2021
The document above follows the rules of the Standard Description for Astronomical Catalogues; from this documentation it is possible to generate f77 program to load files into arrays or line by line