Classifying Pan-STARRS sources with unsupervised and supervised learning#

Learning Goals#

In this tutorial, you will see an example of using models to classify sources extracted from astronomical data. By the end of this tutorial you will have working examples of how to apply principal component analysis (PCA), t-distributed stochastic neighbor embedding (tSNE), k-means clustering (unsupervised learning) and stochastic gradient descent (SGD) classification (supervised learning).

Introduction#

Machine learning which can be applied in both supervised and unsupervised contexts for classification. In this notebook, we will walk through the basic steps of reducing the dimensionality of astronomical source catalogs with PCA and tSNE and classifying the sources in unsupervised and supervised formats.

Download and import the data
Perform data cleaning steps
Reduce dimensionality with PCA and tSNE
Perform unsupervised clustering with k-means
Perform supervised classification with SGD
Assess the success of the model with a confusion matrix

Here we will classify sources extracted from multi-wavelength photometry from the Pan-STARRS survey, and compare with previous classifications from a convolutional neural network analysis by the PS1-STRM team.

Imports#

This notebook uses the following:

numpy to handle array functions
astropy for accessing FITS files
matplotlib.pyplot for plotting data
sklearn for models and performance metrics
itertools for generating combinations
scipy for statistics

If you do not have these packages installed, you can install them using pip or conda.

# arrays
import numpy as np

# fits
from astropy.io import fits
from astropy.utils.data import download_file
from astropy.table import Table

# Plotting
import matplotlib.pyplot as plt

# sklearn for performance metrics

# sklearn 
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import train_test_split
from sklearn.cluster import KMeans
from sklearn.manifold import TSNE
from sklearn.decomposition import PCA
from sklearn.linear_model import SGDClassifier
from sklearn.metrics import confusion_matrix
from sklearn import metrics

# iterative combinations
from itertools import combinations

# statistics 
import scipy.stats as st

# from IPython import get_ipython
# get_ipython().run_line_magic('matplotlib', 'notebook')

# set random seed for reproducibility 
np.random.seed(42)

1. Download the data#

For the following exercises, we will use data from the Pan-STARRS1 (PS1) catalog, which includes wide-area imaging in five wavelength bands, g, r, i, z and y.

Using convolutional neural networks (CNNs) trained on spectroscopic surveys, the Pan-STARRS1 Source Types and Redshifts with Machine Learning (PS1-STRM) (Beck et al. 2020) team classified (“star”, “galaxy”, “QSO” and “unsure”) and regressed redshift information for the 2,902,054,648 discree sources in the PS1 3\(\pi\) DR1 catalog. This volume of data is far too large to analyze in a short notebook example. Therefore, we extracted a subset of PS1-STRM and cross-matched it with the PS1 catalog (see Appendix for details on how to obtain this subset from MAST).

This subset of PS1 will be used to test whether simple clustering algorithms can reproduce the classification results of PS1-STRM. The dataset consists of 100,000 sources, randomly selected to have roughly equal representation in the four PS1-STRM classes.

First, load the data into an astropy table:

%%time
file_url = 'https://archive.stsci.edu/hlsps/hellouniverse/hellouniverse_ps1strm_subset.fits'
tab = Table.read(download_file(file_url, cache=True))

CPU times: user 672 ms, sys: 497 ms, total: 1.17 s
Wall time: 6.75 s

Inspect the contents of the table:

tab

Table length=100000

class	prob_Galaxy	prob_Star	prob_QSO	z_phot	z_photErr	objID	uniquePspsFOid	ippObjID	randomForcedObjID	nDetections	batchID	processingVersion	gnTotal	gnIncPSFFlux	gnIncKronFlux	gnIncApFlux	gnIncR5	gnIncR6	gnIncR7	gFPSFFlux	gFPSFFluxErr	gFPSFFluxStd	gFPSFMag	gFPSFMagErr	gFKronFlux	gFKronFluxErr	gFKronFluxStd	gFKronMag	gFKronMagErr	gFApFlux	gFApFluxErr	gFApFluxStd	gFApMag	gFApMagErr	gFmeanflxR5	gFmeanflxR5Err	gFmeanflxR5Std	gFmeanflxR5Fill	gFmeanMagR5	gFmeanMagR5Err	gFmeanflxR6	gFmeanflxR6Err	gFmeanflxR6Std	gFmeanflxR6Fill	gFmeanMagR6	gFmeanMagR6Err	gFmeanflxR7	gFmeanflxR7Err	gFmeanflxR7Std	gFmeanflxR7Fill	gFmeanMagR7	gFmeanMagR7Err	gFlags	gE1	gE2	rnTotal	rnIncPSFFlux	rnIncKronFlux	rnIncApFlux	rnIncR5	rnIncR6	rnIncR7	rFPSFFlux	rFPSFFluxErr	rFPSFFluxStd	rFPSFMag	rFPSFMagErr	rFKronFlux	rFKronFluxErr	rFKronFluxStd	rFKronMag	rFKronMagErr	rFApFlux	rFApFluxErr	rFApFluxStd	rFApMag	rFApMagErr	rFmeanflxR5	rFmeanflxR5Err	rFmeanflxR5Std	rFmeanflxR5Fill	rFmeanMagR5	rFmeanMagR5Err	rFmeanflxR6	rFmeanflxR6Err	rFmeanflxR6Std	rFmeanflxR6Fill	rFmeanMagR6	rFmeanMagR6Err	rFmeanflxR7	rFmeanflxR7Err	rFmeanflxR7Std	rFmeanflxR7Fill	rFmeanMagR7	rFmeanMagR7Err	rFlags	rE1	rE2	inTotal	inIncPSFFlux	inIncKronFlux	inIncApFlux	inIncR5	inIncR6	inIncR7	iFPSFFlux	iFPSFFluxErr	iFPSFFluxStd	iFPSFMag	iFPSFMagErr	iFKronFlux	iFKronFluxErr	iFKronFluxStd	iFKronMag	iFKronMagErr	iFApFlux	iFApFluxErr	iFApFluxStd	iFApMag	iFApMagErr	iFmeanflxR5	iFmeanflxR5Err	iFmeanflxR5Std	iFmeanflxR5Fill	iFmeanMagR5	iFmeanMagR5Err	iFmeanflxR6	iFmeanflxR6Err	iFmeanflxR6Std	iFmeanflxR6Fill	iFmeanMagR6	iFmeanMagR6Err	iFmeanflxR7	iFmeanflxR7Err	iFmeanflxR7Std	iFmeanflxR7Fill	iFmeanMagR7	iFmeanMagR7Err	iFlags	iE1	iE2	znTotal	znIncPSFFlux	znIncKronFlux	znIncApFlux	znIncR5	znIncR6	znIncR7	zFPSFFlux	zFPSFFluxErr	zFPSFFluxStd	zFPSFMag	zFPSFMagErr	zFKronFlux	zFKronFluxErr	zFKronFluxStd	zFKronMag	zFKronMagErr	zFApFlux	zFApFluxErr	zFApFluxStd	zFApMag	zFApMagErr	zFmeanflxR5	zFmeanflxR5Err	zFmeanflxR5Std	zFmeanflxR5Fill	zFmeanMagR5	zFmeanMagR5Err	zFmeanflxR6	zFmeanflxR6Err	zFmeanflxR6Std	zFmeanflxR6Fill	zFmeanMagR6	zFmeanMagR6Err	zFmeanflxR7	zFmeanflxR7Err	zFmeanflxR7Std	zFmeanflxR7Fill	zFmeanMagR7	zFmeanMagR7Err	zFlags	zE1	zE2	ynTotal	ynIncPSFFlux	ynIncKronFlux	ynIncApFlux	ynIncR5	ynIncR6	ynIncR7	yFPSFFlux	yFPSFFluxErr	yFPSFFluxStd	yFPSFMag	yFPSFMagErr	yFKronFlux	yFKronFluxErr	yFKronFluxStd	yFKronMag	yFKronMagErr	yFApFlux	yFApFluxErr	yFApFluxStd	yFApMag	yFApMagErr	yFmeanflxR5	yFmeanflxR5Err	yFmeanflxR5Std	yFmeanflxR5Fill	yFmeanMagR5	yFmeanMagR5Err	yFmeanflxR6	yFmeanflxR6Err	yFmeanflxR6Std	yFmeanflxR6Fill	yFmeanMagR6	yFmeanMagR6Err	yFmeanflxR7	yFmeanflxR7Err	yFmeanflxR7Std	yFmeanflxR7Fill	yFmeanMagR7	yFmeanMagR7Err	yFlags	yE1	yE2
bytes6	float64	float64	float64	float64	float64	int64	int64	int64	float64	int64	int64	int64	int64	int64	int64	int64	int64	int64	int64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	int64	float64	float64	int64	int64	int64	int64	int64	int64	int64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	int64	float64	float64	int64	int64	int64	int64	int64	int64	int64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	int64	float64	float64	int64	int64	int64	int64	int64	int64	int64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	int64	float64	float64	int64	int64	int64	int64	int64	int64	int64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	float64	int64	float64	float64
GALAXY	0.980312	0.004782678	0.014905294	0.471364349126816	4.93913290028734	200402322441266604	3939368000018016	356370616559241	0.703470831460758	83	3939368	3	14	0	0	0	0	0	0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	16416	-999.0	-999.0	20	2	2	2	2	2	2	5.90703984926222e-06	1.7672400431934e-08	0.000323351006954908	21.9715995788574	0.00324800005182624	3.25623987009749e-05	3.41189002028841e-06	2.69455003738403	20.1182994842529	0.113762997090817	1.9500199414324e-05	9.20953979743899e-09	8.78130013006739e-05	20.6749992370605	0.000513000006321818	2.96671005344251e-05	2.79615005638334e-06	4.15236991102574e-06	0.974824011325836	20.2194004058838	0.10233099758625	7.4609401053749e-05	4.38834013039013e-06	8.68179995450191e-07	1.00258994102478	19.2180995941162	0.0638599991798401	0.000167539998074062	7.01975977790426e-06	1.16333003461477e-05	1.00093996524811	18.3397998809814	0.0454909987747669	16416	0.171017006039619	0.130113005638123	21	9	8	8	9	9	9	4.33265995525289e-06	4.05088997013081e-07	0.2674959897995	22.3082008361816	0.101511999964714	5.15503006681683e-06	1.99629994313e-06	2.35437989234924	22.119499206543	0.420453995466232	3.32163995153678e-06	1.73556998106505e-06	6.28458023071289	22.596700668335	0.567304015159607	6.54475979899871e-06	4.31527996624936e-06	1.42750996019458e-05	0.974824011325836	21.8603000640869	0.715879023075104	1.62930991791654e-05	6.75111004966311e-06	3.42612984240986e-05	0.993659019470215	20.8701000213623	0.449878990650177	4.30016989412252e-05	1.0813700100698e-05	8.05731033324264e-05	0.988604009151459	19.8164005279541	0.273032993078232	114720	0.435211986303329	0.452488988637924	15	2	2	2	2	2	2	9.07123023807799e-07	1.30901003103645e-06	0.100748002529144	24.0058994293213	1.5667599439621	2.49327990786696e-06	8.35970968182664e-06	0.218798995018005	22.9081001281738	3.64036011695862	3.70626003132202e-06	3.32830995830591e-06	0.651328027248383	22.4776992797852	0.975014984607697	3.51279005883498e-08	7.84874009696068e-06	7.69306006986881e-06	0.974824011325836	27.5359001159668	242.589996337891	8.15007024357328e-06	1.23233003250789e-05	8.9091499830829e-06	1.00258994102478	21.622200012207	1.64169001579285	-1.1616200026765e-05	1.97320005099755e-05	8.54328027344309e-06	1.00093996524811	-999.0	-999.0	16416	13.102499961853	5.33730983734131	13	3	3	3	3	3	3	6.04524984737509e-06	5.89502997172531e-06	1.01250004768372	21.9465007781982	1.05876004695892	1.14544000098249e-05	2.82598994090222e-05	4.9345498085022	21.2525997161865	2.67868995666504	2.03251001948956e-05	1.01291998362285e-05	5.01808977127075	20.6299991607666	0.54108601808548	1.49158004205674e-05	1.5456700566574e-05	2.05233991437126e-05	0.974824011325836	20.9659004211426	1.12511003017426	2.12734994420316e-05	2.42331007029861e-05	2.80195999948774e-05	1.00258994102478	20.5804996490479	1.23678994178772	3.41193008353002e-05	3.88197986467276e-05	9.14734991965815e-05	1.00093996524811	20.0676002502441	1.23531997203827	16416	-0.532278001308441	-0.118960998952389
QSO	0.03350833	0.068698995	0.89779264	-999.0	-999.0	200402338755108032	3940032000001551	356374911404730	0.150013838103054	122	3940032	3	17	9	9	8	9	9	9	1.19898995762924e-05	8.32552984775248e-07	1.67218005657196	21.2029991149902	0.0753910019993782	1.07902997115161e-05	1.10330995539698e-06	2.48463010787964	21.3174991607666	0.111017003655434	1.17192003017408e-05	9.62101012191852e-07	4.22345018386841	21.2278003692627	0.0891349986195564	1.17475001388812e-05	2.75965999207983e-06	2.75726006293553e-06	0.974824011325836	21.2252006530762	0.255055010318756	1.10003002191661e-05	4.29943020208157e-06	5.79853985982481e-06	1.00258994102478	21.2966003417969	0.424358993768692	8.71554038894828e-06	6.86453995513148e-06	9.61031037149951e-06	1.00038003921509	21.5492992401123	0.85514897108078	114992	-0.0240768007934093	-0.0271390005946159	31	16	16	15	16	16	16	1.27534003695473e-05	5.77301022985921e-07	1.26760005950928	21.1359996795654	0.0491469986736774	1.26545000966871e-05	6.5755801870182e-07	1.26300001144409	21.1445007324219	0.0564169995486736	1.40440997711266e-05	6.84298981923348e-07	2.06064009666443	21.0312995910645	0.0529030002653599	1.27162002172554e-05	3.79860989596637e-06	3.76233992938069e-06	0.950923025608063	21.1392002105713	0.324333995580673	1.21768998724292e-05	5.90783020015806e-06	1.09892998807481e-05	0.956041991710663	21.1861991882324	0.526763021945953	9.03611999092391e-06	9.38420998863876e-06	1.97494991880376e-05	0.924996018409729	21.5100994110107	1.1275600194931	115000	-0.0406830012798309	0.0131635004654527	36	24	24	23	24	24	24	1.4342100257636e-05	4.83518022065255e-07	0.848752975463867	21.0084991455078	0.0366039983928204	1.45406002047821e-05	1.27224996049335e-06	2.37408995628357	20.9936008453369	0.0949980020523071	1.65858000400476e-05	9.97104962152662e-07	5.89462995529175	20.850700378418	0.0652720034122467	1.50718997247168e-05	4.49143999503576e-06	6.14887994743185e-06	0.967916011810303	20.9545993804932	0.32354998588562	1.79282997123664e-05	6.98812982591335e-06	1.07981995824957e-05	0.976988017559052	20.7661991119385	0.423200994729996	2.69341999228345e-05	1.09785996755818e-05	2.39020992012229e-05	0.928957998752594	20.3243007659912	0.442555993795395	115000	-0.0434597991406918	0.00493281008675694	21	16	16	14	16	16	16	1.65381006809184e-05	9.50451010339748e-07	0.871819972991943	20.8539009094238	0.0623980015516281	1.60365998453926e-05	1.75684999703662e-06	2.07728004455566	20.887300491333	0.118945002555847	2.01734001166187e-05	1.31707997752528e-06	3.34904003143311	20.6380996704102	0.0708860009908676	1.81292998604476e-05	7.02020997778163e-06	1.25544002003153e-05	0.974824011325836	20.7541007995605	0.420430988073349	2.43525992118521e-05	1.09956999949645e-05	2.64229001913918e-05	1.0011500120163	20.4337005615234	0.490233987569809	2.21015998249641e-05	1.75948007381521e-05	5.80384003114887e-05	0.99966299533844	20.5389995574951	0.864340007305145	114720	-0.182536005973816	0.329957991838455	17	8	8	8	8	8	8	1.62028991326224e-05	3.72488989341946e-06	0.590462982654572	20.8761005401611	0.249601006507874	3.92274996556807e-05	1.03520997072337e-05	1.97853004932404	19.9160995483398	0.286523997783661	2.89184008579468e-05	9.24110008782009e-06	5.47257995605469	20.2471008300781	0.346955001354218	2.34809995163232e-05	1.9519500710885e-05	2.56544008152559e-05	0.974824011325836	20.4733009338379	0.902558982372284	4.16576986026485e-05	3.05850990116596e-05	4.85365017084405e-05	1.00258994102478	19.8507995605469	0.79714697599411	6.60782970953733e-05	4.89935991936363e-05	0.000138739997055382	1.00093996524811	19.3498992919922	0.805015981197357	16416	1.54497003555298	-0.447537988424301
UNSURE	0.5863613	0.027802624	0.38583606	-999.0	-999.0	200402414126263504	3941360000008952	356520940352606	0.0540886456650576	95	3941360	3	13	3	3	2	3	3	3	3.95490997107117e-06	7.74327020280907e-07	0.25499901175499	22.4071998596191	0.212575003504753	6.31689999863738e-06	2.89971990241611e-06	3.46514010429382	21.8987998962402	0.498398989439011	8.73719000082929e-06	2.06251002055069e-06	3.28488993644714	21.5466003417969	0.256300002336502	5.87131989959744e-06	2.50342009167071e-06	3.87454019801226e-06	0.905561983585358	21.9782009124756	0.462938010692596	6.75093997415388e-06	3.77722994926444e-06	5.02656985190697e-06	0.843654990196228	21.8267002105713	0.607482016086578	5.84400004299823e-06	5.62508012080798e-06	1.93601008504629e-05	0.732827007770538	21.9832992553711	1.04507005214691	16416	-0.397132992744446	0.524013996124268	23	8	8	8	8	8	8	1.89089996638359e-06	5.37540017830906e-07	0.987200021743774	23.2084007263184	0.308649986982346	8.52970970299793e-06	3.65708001481835e-06	5.54893016815186	21.5727005004883	0.465505003929138	7.84334042691626e-06	3.15343004331226e-06	23.4568996429443	21.6637992858887	0.436522990465164	5.39100983587559e-06	3.2218099477177e-06	6.6306001826888e-06	0.974824011325836	22.0708999633789	0.648863971233368	9.96721973933745e-06	5.05231992065092e-06	1.30841999634868e-05	1.00248003005981	21.403600692749	0.550352990627289	1.87213008757681e-05	8.02893009677064e-06	2.43733993556816e-05	0.969539999961853	20.7192001342773	0.465635985136032	16416	0.36225301027298	-0.0884943976998329	30	11	10	8	11	11	11	2.73685009233304e-06	8.94921981853258e-07	1.57816994190216	22.8069000244141	0.355024993419647	1.84915995760093e-06	6.99815984717134e-07	1.96721994876862	23.2325992584229	0.410899013280869	2.39618998421065e-06	6.71243014949141e-07	1.98375999927521	22.9512996673584	0.304145991802216	2.07477000913059e-06	4.53353004559176e-06	1.65962992468849e-05	0.900672018527985	23.1075992584229	2.37242007255554	9.29829002416227e-06	6.96289998813882e-06	3.41589984600432e-05	0.908926010131836	21.4790992736816	0.813039004802704	1.81426003109664e-05	1.10108003354981e-05	5.9927100664936e-05	0.901485979557037	20.7532997131348	0.658936977386475	114720	-0.169953003525734	-0.140532001852989	14	2	2	2	2	2	2	1.26121003631852e-05	3.39051007358648e-06	0.855762004852295	21.148099899292	0.291878998279572	-2.07172997761518e-05	1.89740003406769e-05	1.87910997867584	-999.0	-999.0	-2.5691199425637e-06	2.62675007434154e-06	0.513532996177673	-999.0	-999.0	-3.39366010848607e-06	8.2760698205675e-06	6.11017003393499e-06	0.872036993503571	-999.0	-999.0	-1.12812003862928e-05	1.25260003187577e-05	5.46997989658848e-06	0.846889019012451	-999.0	-999.0	-1.47434002428781e-05	1.95359007193474e-05	3.57741992047522e-05	0.810638010501862	-999.0	-999.0	16416	-0.246417999267578	-0.0802292004227638	15	2	2	2	2	2	2	2.40323006437393e-05	2.59417993220268e-06	0.0608064010739326	20.4480991363525	0.117200002074242	4.53460015705787e-05	7.25369027350098e-05	15.1094999313354	19.7586994171143	1.73678004741669	6.24618987785652e-05	6.49142966722138e-05	38.1203994750977	19.4109992980957	1.1283700466156	4.43028984591365e-05	2.21804002649151e-05	6.70418012305163e-05	0.974824011325836	19.7840003967285	0.54357898235321	0.000138413000968285	3.48079011018854e-05	0.000193526997463778	1.00258994102478	18.5471000671387	0.273039996623993	0.000272422999842092	5.57392995688133e-05	0.000348463014233857	1.00093996524811	17.8120002746582	0.222148001194	16416	0.0850384011864662	0.265985012054443
GALAXY	0.95117044	0.008774751	0.040054727	0.664705097675323	0.179729584695471	200412343261577739	3940032000013316	356374911539352	0.070737426235096	122	3940032	3	17	3	3	3	3	3	3	1.64772995958629e-06	5.62291006644955e-07	0.985023021697998	23.3577995300293	0.37050798535347	1.55965994963481e-06	8.51050003802811e-07	1.35141003131866	23.4174995422363	0.592446029186249	1.84554005500104e-06	9.37632023578772e-07	6.00352001190186	23.2348003387451	0.551609992980957	1.2765300425599e-06	1.76149001163139e-06	1.51047004237626e-06	0.974824011325836	23.6350002288818	1.49820995330811	4.77353012229287e-07	2.790769940475e-06	2.64317009168735e-06	1.00258994102478	24.7029991149902	6.34758996963501	9.78651996774715e-07	4.51463984063594e-06	1.44173000080627e-05	1.00093996524811	23.9235000610352	5.00863981246948	16416	-0.649612009525299	0.371650010347366	31	7	7	5	7	7	7	3.40854990099615e-06	7.21119022273342e-07	1.01329004764557	22.5685997009277	0.229699999094009	7.92841001384659e-06	2.31601006817073e-06	3.85576009750366	21.652099609375	0.31716001033783	7.92528044257779e-06	1.2505299764598e-06	3.98901009559631	21.6525001525879	0.171317994594574	6.90432989358669e-06	3.48902995028766e-06	7.77954028308159e-06	0.961561024188995	21.8022994995117	0.54866498708725	6.14849977864651e-06	5.35109984411974e-06	1.34620004246244e-05	0.971174001693726	21.9281005859375	0.944926023483276	6.61755984765477e-06	8.39958011056297e-06	2.14045994653134e-05	0.952320992946625	21.8483009338379	1.37811005115509	16416	0.358875006437302	-1.45696997642517	36	16	15	15	16	16	16	3.2864199965843e-06	3.14049003691252e-07	0.429374009370804	22.6082992553711	0.103753000497818	2.73110003945476e-06	1.4907000149833e-06	1.96333003044128	22.8092002868652	0.592620015144348	3.57653993887652e-06	1.22302003546793e-06	5.35037994384766	22.5163993835449	0.371273994445801	2.35009997595625e-06	4.36667005487834e-06	6.25667007625452e-06	0.955344021320343	22.9724006652832	2.01737999916077	-2.6615998649504e-07	6.77060006637475e-06	1.16730998342973e-05	0.958121001720428	-999.0	-999.0	-7.57049019739497e-06	1.0693500371417e-05	2.68048006546451e-05	0.926276981830597	-999.0	-999.0	114720	-1.0854799747467	-0.934072017669678	21	8	8	6	8	8	8	4.69525002699811e-06	7.47454976135487e-07	0.539090991020203	22.2208995819092	0.172842994332314	9.46860018302687e-06	4.22798984800465e-06	3.10921001434326	21.459400177002	0.484811007976532	5.3207600103633e-06	1.7358199784212e-06	1.57735002040863	22.0851001739502	0.354205012321472	9.00533996173181e-06	6.72818987368373e-06	1.0941100299533e-05	0.868996977806091	21.5137996673584	0.811190009117126	1.94339991139714e-05	1.00805000329274e-05	2.14970004890347e-05	0.843770980834961	20.6786994934082	0.563176989555359	3.01502004731447e-05	1.54851004481316e-05	3.94499984395225e-05	0.791230022907257	20.2017993927002	0.557635009288788	16416	0.653178989887238	-1.86926996707916	17	3	3	3	3	3	3	2.36168998526409e-05	7.3522501224943e-06	0.972268998622894	20.4669990539551	0.338003993034363	6.96718998369761e-05	6.44177998765372e-06	0.224421992897987	19.2924003601074	0.100386001169682	7.30633983039297e-05	5.01224985782756e-06	0.595785975456238	19.2408008575439	0.0744829997420311	5.58997016923968e-05	2.11403003049782e-05	3.17623998853378e-05	0.974824011325836	19.5314998626709	0.410605996847153	9.13265976123512e-05	3.32076015183702e-05	7.12602995918132e-05	1.00258994102478	18.9986000061035	0.394789010286331	0.00016030200640671	5.30573015566915e-05	0.000250007986323908	0.986687004566193	18.3876991271973	0.359360009431839	16416	0.119231000542641	0.136751994490623
STAR	0.007560004	0.990866	0.0015739122	-999.0	-999.0	200412436637435074	3941763000002733	356525235265566	0.73679818747267	230	3941763	3	26	9	9	7	9	9	9	3.12768997901003e-06	4.20390989575026e-07	0.820297002792358	22.6620006561279	0.145933002233505	4.98919007441145e-06	1.04928994915099e-06	3.13581991195679	22.1550006866455	0.228342995047569	5.1218198677816e-06	9.23576976674667e-07	5.28777980804443	22.1264991760254	0.195782005786896	4.67217978439294e-06	2.57391002378426e-06	4.79267009723117e-06	0.970157027244568	22.2262992858887	0.598133027553558	6.43055000182358e-06	3.99606005885289e-06	9.15596956474474e-06	0.978361010551453	21.8794002532959	0.674696981906891	1.07744999695569e-05	6.19254979028483e-06	1.82240000867751e-05	0.904103994369507	21.3190994262695	0.624014973640442	114720	1.24248003959656	3.19806003570557	48	19	19	19	19	19	19	7.85234988143202e-06	4.01471993427549e-07	0.87984299659729	21.6625995635986	0.0555110014975071	6.09513017479912e-06	5.9397797258498e-07	1.6355299949646	21.9375991821289	0.10580600053072	6.48624018140254e-06	8.13708027180837e-07	4.96670007705688	21.8701000213623	0.13620699942112	6.88597992848372e-06	4.71617022412829e-06	6.36859022051794e-06	0.974591016769409	21.8052005767822	0.743614017963409	7.95211963122711e-06	7.3815699579427e-06	8.22572019387735e-06	0.997386991977692	21.6488990783691	1.00784003734589	1.44333998832735e-05	1.18105999717955e-05	2.51574001595145e-05	0.988421022891998	21.0016002655029	0.888442993164063	115000	0.0858049988746643	0.0611580014228821	76	29	28	27	29	29	29	2.07773991860449e-05	5.72544024635135e-07	1.4680700302124	20.6061000823975	0.0299190003424883	1.65649998962181e-05	7.42256986541179e-07	1.40615999698639	20.8521003723145	0.048650998622179	1.72613999893656e-05	1.05151002571802e-06	5.72194004058838	20.8073997497559	0.0661400035023689	1.65110996022122e-05	5.97044981986983e-06	8.87021997186821e-06	0.974214971065521	20.8556003570557	0.392603993415833	1.42512999445898e-05	9.32900002226233e-06	1.9054099539062e-05	0.990769982337952	21.0153999328613	0.710731029510498	4.92194021717296e-06	1.4804400052526e-05	4.79127011203673e-05	0.956910014152527	22.1697006225586	3.26571989059448	115000	-0.0426842011511326	0.0197713002562523	38	16	16	14	16	16	16	3.48864014085848e-05	9.92064997262787e-07	1.48014998435974	20.043399810791	0.0308749992400408	3.42837993230205e-05	1.84737996278272e-06	1.96736001968384	20.0622997283936	0.0585049986839294	3.82176986022387e-05	1.49626998791064e-06	4.02335977554321	19.9444007873535	0.0425079986453056	3.59564983227756e-05	7.31821000954369e-06	1.17846002467559e-05	0.951061010360718	20.0105991363525	0.220979005098343	4.1801598854363e-05	1.13491996671655e-05	2.34296003327472e-05	0.947552978992462	19.8470993041992	0.294779002666473	5.732950012316e-05	1.79527996806428e-05	5.91530988458544e-05	0.920458972454071	19.5041007995605	0.340000003576279	115000	-0.00746937980875373	0.0482023991644382	42	14	15	11	15	15	15	4.21131007897202e-05	2.34761000683648e-06	1.52991998195648	19.8390007019043	0.0605250000953674	3.71136993635446e-05	4.87347006128402e-06	3.01684999465942	19.9762001037598	0.142570003867149	4.07447987527121e-05	4.77144021715503e-06	4.37614011764526	19.8749008178711	0.127146005630493	4.04083984903991e-05	1.89040001714602e-05	3.42556995747145e-05	0.974824011325836	19.8838996887207	0.507933020591736	5.52484998479486e-05	2.95614991046023e-05	7.76672022766434e-05	0.998709976673126	19.5443000793457	0.58093798160553	8.07286996860057e-05	4.73750988021493e-05	0.000173113003256731	0.998739004135132	19.1324996948242	0.63715797662735	115000	-0.148842006921768	-0.0163633991032839
GALAXY	0.9514053	0.033890028	0.014704614	0.701190054416656	0.0719254109262685	200412465551166728	3942332000003895	356529530259173	0.666663128572202	130	3942332	3	14	9	8	5	9	9	9	1.46972001857648e-06	2.9355499009398e-07	0.508948981761932	23.4820003509521	0.216859996318817	3.08114999825193e-06	9.09450022845704e-07	2.26735997200012	22.6783008575439	0.320472002029419	3.48392995874747e-06	3.79961988983268e-07	0.386496007442474	22.544900894165	0.118412002921104	3.38412996825355e-06	2.2608001017943e-06	3.21897005051142e-06	0.974824011325836	22.5764007568359	0.725335001945496	6.20090986558353e-06	3.54087001142034e-06	8.23985010356409e-06	1.00258994102478	21.9188995361328	0.619982004165649	9.74189970293082e-06	5.61990009373403e-06	1.62885007739533e-05	0.982533991336823	21.4284992218018	0.626338005065918	16416	-0.631915986537933	0.158441007137299	34	12	11	12	12	12	12	2.81646998701035e-06	2.40849999499915e-07	0.405898988246918	22.7758007049561	0.0928469970822334	2.4399100766459e-06	7.56677025037789e-07	1.31983995437622	22.9316005706787	0.336715012788773	3.76899993170809e-06	1.08654000996466e-06	5.16838979721069	22.4594993591309	0.31299901008606	3.00323995361396e-06	2.97626002065954e-06	3.35820004693232e-06	0.956956028938293	22.7061004638672	1.07597994804382	3.17918988912425e-06	4.68869984615594e-06	5.48132993571926e-06	0.989808976650238	22.6443004608154	1.60125005245209	7.13288000042667e-06	7.52983987695188e-06	1.60281997523271e-05	0.990760028362274	21.7668991088867	1.14616000652313	114720	0.54126900434494	-0.227962002158165	40	16	16	16	16	16	16	7.09222013028921e-06	7.25040990801062e-07	1.30656003952026	21.773099899292	0.110995002090931	6.79279992255033e-06	1.27140003769455e-06	1.76215994358063	21.8199005126953	0.203216001391411	7.46246996641275e-06	1.5216500059978e-06	5.97206020355225	21.7178993225098	0.22138899564743	8.18017997517018e-06	5.82032998863724e-06	9.42506994761061e-06	0.974686026573181	21.618200302124	0.772518992424011	7.83102041168604e-06	9.12031009647762e-06	1.18135003503994e-05	1.00021994113922	21.6655006408691	1.26449000835419	-5.03428009324125e-06	1.4604500393034e-05	3.45814005413558e-05	0.997327029705048	-999.0	-999.0	114992	0.893664002418518	0.161718994379044	20	7	7	6	7	7	7	7.45192983231391e-06	1.09346001408994e-06	0.617147982120514	21.7194004058838	0.159316003322601	6.06653020440717e-06	5.07778986502672e-06	3.75811004638672	21.942699432373	0.90877902507782	3.42524003826838e-06	4.02361001761165e-06	6.53002977371216	22.5632991790771	1.2754100561142	6.87381998432102e-06	7.46596015233081e-06	1.44072000693996e-05	0.973877012729645	21.8071002960205	1.17927002906799	8.05224044597708e-06	1.15479997475632e-05	2.58254003711045e-05	0.981602013111115	21.6352996826172	1.55709004402161	-6.74741977491067e-06	1.79254002432572e-05	4.41898991994094e-05	0.939588010311127	-999.0	-999.0	118840	0.155974999070168	-1.2666699886322	22	7	7	5	7	7	7	1.24837997645955e-05	3.00384999718517e-06	0.599393010139465	21.159200668335	0.261247992515564	1.82289004442282e-05	1.28304000099888e-05	6.22979021072388	20.7481994628906	0.764192998409271	1.83126994670602e-05	3.64651009476802e-06	1.45959997177124	20.743200302124	0.216196998953819	2.15268992178608e-05	2.11023998417659e-05	5.44643007742707e-05	0.969139993190765	20.5676002502441	1.06432998180389	6.27726012680796e-06	3.27599009324331e-05	5.15659994562157e-05	0.976849973201752	21.9055995941162	5.66626977920532	-2.71238004643237e-05	5.20167996000964e-05	9.76180017460138e-05	0.96120297908783	-999.0	-999.0	16416	0.415657997131348	-0.363719999790192
GALAXY	0.79322743	0.17576069	0.031011794	0.677324831485748	0.244630052392452	200422297094590056	3938806000009635	356366321560210	0.575081246509292	82	3938806	3	11	3	3	3	3	3	3	8.17628972527018e-07	3.74233991351502e-07	0.40489798784256	24.1187000274658	0.496948003768921	5.00710984852049e-06	1.24618998142978e-06	0.67280501127243	22.1511001586914	0.270224004983902	3.95864981328486e-06	1.42314002005151e-06	3.85121989250183	22.4062004089355	0.390323996543884	3.49881997863122e-06	2.18756008507626e-06	2.41187990468461e-06	0.950509011745453	22.5403003692627	0.678834021091461	5.06859987581265e-06	3.4214899642393e-06	3.48208004652406e-06	0.973330020904541	22.1378002166748	0.732910990715027	1.08306003312464e-05	5.31427986061317e-06	1.09177999547683e-05	0.895157992839813	21.3134002685547	0.532739996910095	16416	-0.288262009620667	0.666837990283966	20	5	5	4	5	5	5	2.28841008720337e-06	4.94957021146547e-07	0.621154010295868	23.00119972229	0.234832003712654	1.48455001180992e-06	1.75465004303987e-06	1.99085998535156	23.4710998535156	1.28328001499176	1.43893998938438e-06	9.47149999319663e-07	1.29446995258331	23.5049991607666	0.714663028717041	1.53025996496581e-06	3.01931004287326e-06	3.98289012082387e-06	0.974824011325836	23.4381999969482	2.14223003387451	2.26742002951141e-07	4.73025011160644e-06	8.51312961458461e-06	1.00258994102478	25.5111999511719	22.6504993438721	1.50988000768848e-06	7.55585006118054e-06	2.22758008021628e-05	0.989441990852356	23.4526996612549	5.43330001831055	16416	-0.291516005992889	-0.154441997408867	21	4	3	2	4	4	4	4.22301991420682e-06	3.56664008904772e-06	2.94200992584229	22.3360004425049	0.916980981826782	3.25818982673809e-05	7.89663044997724e-06	8.36746978759766	20.1175994873047	0.263141989707947	3.79405992134707e-06	4.58911017631181e-06	3.91392993927002	22.452299118042	1.31324994564056	7.41933990866528e-06	4.96676011607633e-06	1.21336997835897e-05	0.974824011325836	21.7241992950439	0.726828992366791	2.46378003794234e-05	7.80076970841037e-06	2.75856000371277e-05	1.00258994102478	20.4211006164551	0.343764007091522	4.33729983342346e-05	1.24948001030134e-05	5.4813801398268e-05	1.0008499622345	19.806999206543	0.312775999307632	16416	-0.154929995536804	-0.0871376022696495	14	2	2	2	2	2	2	5.82117024805484e-07	1.14639999537758e-06	0.178731992840767	24.4874992370605	2.13821005821228	1.93905998457922e-05	7.16155000191065e-07	0.00504788989201188	20.6811008453369	0.0401000007987022	1.10721002783976e-05	8.7896302147783e-07	0.105062000453472	21.2894992828369	0.0861919969320297	1.14140002551721e-05	6.66261985315941e-06	1.57299000420608e-05	0.974824011325836	21.2565002441406	0.633772015571594	1.45024996527354e-05	1.04852997537819e-05	8.36017989058746e-06	1.00258994102478	20.9965000152588	0.784990012645721	-1.192780018755e-05	1.66262998391176e-05	4.1579198295949e-05	0.984950006008148	-999.0	-999.0	16416	-18.0902004241943	37.8647003173828	16	6	5	5	6	6	6	1.9562799934647e-05	7.69366033637198e-06	4.8617000579834	20.6714992523193	0.426997989416122	3.01298005069839e-05	1.92253992281621e-05	7.78587007522583	20.202600479126	0.692794024944305	3.36242010234855e-05	1.31585002236534e-05	7.80966997146606	20.0834007263184	0.424892991781235	7.31550971977413e-05	2.25443000090308e-05	7.86886012065224e-05	0.972981989383698	19.2395000457764	0.334592998027802	0.000154919995111413	3.54039002559148e-05	0.000208625002414919	1.00074005126953	18.4248008728027	0.248123005032539	0.000387306994525716	5.67091010452714e-05	0.000479506998090073	0.99806797504425	17.4298992156982	0.15897199511528	115000	0.333539992570877	-0.605168998241425
GALAXY	0.99721175	5.262087e-05	0.0027355852	0.0950550884008408	0.00796880137784138	200422302783812138	3939368000000719	356370616426218	0.987247876811072	166	3939368	3	25	7	11	6	11	11	11	0.000121566001325846	1.78244999915478e-06	8.42679023742676	18.6879997253418	0.0159189999103546	0.000235662999330089	3.28704004459723e-06	6.38078022003174	17.9692993164063	0.0151439998298883	0.000201386006665416	7.26288021724031e-07	1.5615199804306	18.1399993896484	0.00391600001603365	0.000179074995685369	3.05216008200659e-06	2.07353004952893e-05	0.973617970943451	18.2674999237061	0.0185049995779991	0.000223270006245002	4.49162007498671e-06	1.29454001580598e-05	0.999734997749329	18.0279998779297	0.021841999143362	0.000251040008151904	6.82997006151709e-06	1.96618002519244e-05	0.981078028678894	17.9006996154785	0.0295390002429485	16892216	0.0159204006195068	0.0417909994721413	40	7	11	7	12	12	12	0.000235137005802244	3.32867989527585e-06	8.34934997558594	17.9717998504639	0.01537000015378	0.000511545978952199	4.84365000374964e-06	2.97360992431641	17.1278991699219	0.0102800000458956	0.000409947009757161	2.08799997380993e-06	6.68710994720459	17.368200302124	0.00553000019863248	0.000386250991141424	6.18750982539495e-06	3.45266016665846e-05	0.974824011325836	17.4328994750977	0.0173930004239082	0.000483545998577029	9.44254043133697e-06	1.7562300854479e-05	0.998355984687805	17.1889991760254	0.0212019998580217	0.000553885998670012	1.49208999573602e-05	2.56868006545119e-05	0.998127996921539	17.0415000915527	0.0292479991912842	16892216	0.0211217999458313	0.0374512001872063	42	10	17	12	18	18	18	0.000350957008777186	5.16001000505639e-06	21.4904003143311	17.5368995666504	0.0159629993140697	0.000766294018831104	5.63628009331296e-06	4.80072021484375	16.6891002655029	0.00798599980771542	0.000614051998127252	6.32956016488606e-06	46.8216018676758	16.9295997619629	0.0111919995397329	0.000587219023145735	7.14992984285345e-06	3.34785981976893e-05	0.969789028167725	16.978099822998	0.0132200000807643	0.000717920018360019	1.08357999124564e-05	2.48866999754682e-05	0.998175978660584	16.7598991394043	0.0163870006799698	0.000801234971731901	1.69997001648881e-05	4.36543014075141e-05	0.997907996177673	16.6406993865967	0.0230359993875027	16892216	-0.00922395009547472	0.0406208001077175	28	4	6	4	6	6	6	0.000475412991363555	9.48391971178353e-06	17.317699432373	17.2073993682861	0.0216589998453856	0.000973592977970839	1.26612003441551e-05	6.6236400604248	16.4291000366211	0.0141200004145503	0.000796727021224797	4.80060998597764e-06	4.80774021148682	16.6467990875244	0.00654199998825789	0.000760954979341477	7.36334004614037e-06	2.99966995953582e-05	0.954192996025085	16.6966991424561	0.0105060003697872	0.000928746012505144	1.08648000605172e-05	3.67499997082632e-05	0.986728012561798	16.4803009033203	0.0127010000869632	0.00103086000308394	1.65917008416727e-05	4.93838997499552e-05	0.970034003257751	16.3670997619629	0.0174749996513128	16892216	-0.00811044033616781	0.0493528991937637	31	9	10	7	10	10	10	0.000561578024644405	2.41697998717427e-05	43.9789009094238	17.0265007019043	0.0467289984226227	0.00106373999733478	1.3959799616714e-05	4.37357997894287	16.3330001831055	0.0142480004578829	0.0010175199713558	1.68798997037811e-05	12.4666996002197	16.3812007904053	0.0180120002478361	0.000858751009218395	1.84038999577751e-05	8.82847016328014e-05	0.968414008617401	16.5653991699219	0.02326799929142	0.0010701100109145	2.80862004728988e-05	7.58941969252191e-05	0.993997991085052	16.3264999389648	0.0284960009157658	0.00119692995212972	4.44299985247198e-05	0.00011120500130346	0.997600972652435	16.2049007415771	0.040302000939846	16892216	0.0158769004046917	0.0384196005761623
QSO	0.03287622	0.04926253	0.9178612	-999.0	-999.0	200422337243332337	3940032000001091	356374911398983	0.25860480522607	122	3940032	3	17	8	8	5	8	8	8	1.30348998936825e-05	6.93898982717656e-07	2.62853002548218	21.1123008728027	0.0577979981899261	1.40196998472675e-05	1.63945003350818e-06	6.80090999603271	21.0331993103027	0.126965001225472	1.42026001412887e-05	1.7163999928016e-06	9.42453002929688	21.019100189209	0.131211996078491	1.46319998748368e-05	2.77671006188029e-06	8.50950982567156e-06	0.962665975093842	20.9867992401123	0.206039994955063	2.05792002816452e-05	4.35238007412408e-06	1.35126001623576e-05	0.993234992027283	20.6165008544922	0.229626998305321	2.78709994745441e-05	6.95724020260968e-06	1.44217001434299e-05	0.98598301410675	20.2872009277344	0.271025002002716	115000	-0.0423665009438992	0.0733482986688614	31	18	18	17	18	18	18	1.42095996125136e-05	5.00810017456388e-07	1.89259004592896	21.0186004638672	0.0382659994065762	1.47738001032849e-05	1.24640996546077e-06	2.62898993492126	20.9762992858887	0.0916000008583069	1.48862000060035e-05	1.14316003418935e-06	7.76796007156372	20.9680995941162	0.0833770036697388	1.47625996760326e-05	3.77555011255026e-06	5.98887982050655e-06	0.974824011325836	20.9771995544434	0.2776780128479	1.57538997882511e-05	5.90916988585377e-06	1.07875002868241e-05	1.00258994102478	20.9066009521484	0.407251000404358	1.95171996892896e-05	9.46578984439839e-06	3.40115002472885e-05	1.00093996524811	20.673999786377	0.526579976081848	115000	0.0264203008264303	-0.111978001892567	36	14	14	13	14	14	14	1.70755993167404e-05	4.39030998222734e-07	0.790808975696564	20.8190994262695	0.0279150009155273	1.47932996696909e-05	6.27519000317989e-07	0.970220029354095	20.9748992919922	0.0460559986531734	1.55423003889155e-05	1.07388996184454e-06	4.97104978561401	20.9213008880615	0.0750190019607544	1.57094000314828e-05	4.61450008515385e-06	5.3736998779641e-06	0.974824011325836	20.9097003936768	0.318924993276596	1.77919991983799e-05	7.22223012417089e-06	1.0368100447522e-05	1.00021994113922	20.7744998931885	0.44072899222374	2.18110999412602e-05	1.1513599929458e-05	4.0244198316941e-05	0.983896970748901	20.5534000396729	0.573132991790771	115000	0.0628589987754822	0.11183799803257	21	15	15	13	15	15	15	1.5854700905038e-05	1.09003997295076e-06	1.44456994533539	20.8997001647949	0.0746470019221306	1.37112001539208e-05	1.30018997879233e-06	1.04270994663239	21.0573997497559	0.102957002818584	1.89194997801678e-05	1.89421996310557e-06	4.99010992050171	20.707799911499	0.108704000711441	1.52269003592664e-05	7.98933979240246e-06	1.22333003673702e-05	0.969814002513886	20.9435005187988	0.569669008255005	2.85736005025683e-05	1.25254000522546e-05	3.01456002489431e-05	0.995999991893768	20.2602005004883	0.475937992334366	6.44745014142245e-05	2.00654994841898e-05	7.31866966816597e-05	0.995048999786377	19.3766002655029	0.33789798617363	115000	-0.607792019844055	0.0594804994761944	17	9	9	8	9	9	9	2.76549999398412e-05	4.58896010968601e-06	1.78558003902435	20.2956008911133	0.180161997675896	2.33826995099662e-05	4.44785018771654e-06	0.97408801317215	20.4778003692627	0.206529006361961	2.70261007244699e-05	2.90734010377491e-06	0.612289011478424	20.3206005096436	0.116798996925354	3.22277992381714e-05	1.79978997039143e-05	2.07067005248973e-05	0.974824011325836	20.1294994354248	0.606338977813721	7.04771009623073e-05	2.82410001091193e-05	4.55566005257424e-05	1.00258994102478	19.279899597168	0.435066998004913	0.000114845999632962	4.49924009444658e-05	0.000127330000395887	0.986647009849548	18.7497997283936	0.42535001039505	114720	0.363135010004044	-0.21280300617218
QSO	0.043625608	0.050498046	0.90587634	-999.0	-999.0	200400164900318067	4028334000006867	351650447395608	0.680476882197407	87	4028334	3	14	4	4	3	4	4	4	3.91004004995921e-06	1.21209995995741e-06	2.87283992767334	22.4195995330811	0.336575001478195	5.41722010893864e-06	3.26947997564275e-06	6.20335006713867	22.0655994415283	0.655278027057648	4.20423020841554e-06	1.20975005302171e-06	4.72206020355225	22.340799331665	0.31241500377655	6.55122994430712e-06	2.38609004554746e-06	6.56813017485547e-06	0.967086970806122	21.8593006134033	0.395446985960007	8.24397011456313e-06	3.75641002392513e-06	9.01903968042461e-06	0.996585011482239	21.6096992492676	0.494720995426178	9.8035998234991e-06	6.06512003287207e-06	2.06001004698919e-05	0.990251004695892	21.4216003417969	0.671703994274139	16793912	0.260302007198334	0.00549450982362032	16	10	10	10	10	10	10	6.57533018966205e-06	7.43670000247221e-07	1.21698999404907	21.8553009033203	0.12279699742794	5.77187984163174e-06	1.85332999080856e-06	4.24852991104126	21.9967994689941	0.348625987768173	6.61889998809784e-06	1.65481003477908e-06	10.5132999420166	21.8481006622314	0.271448999643326	4.49400022262125e-06	4.64389995613601e-06	9.84869984677061e-06	0.942551016807556	22.2684993743896	1.12195003032684	8.33436024549883e-06	6.84195993017056e-06	1.5638299373677e-05	0.910278975963593	21.597900390625	0.891318023204803	5.21309993928298e-06	1.02632002381142e-05	2.57872998190578e-05	0.840038001537323	22.1072998046875	2.13753008842468	114720	1.44615995883942	1.31737995147705	18	9	9	7	9	9	9	7.2514999374107e-06	7.01021974691685e-07	0.778777003288269	21.7490005493164	0.1049610003829	6.21892013441538e-06	2.44538000515604e-06	1.76130998134613	21.9158000946045	0.426928997039795	5.656099801854e-06	1.01450996226049e-06	2.0521399974823	22.0188007354736	0.194743007421494	4.02917021347093e-06	4.74832995678298e-06	7.24728988643619e-06	0.956157982349396	22.386999130249	1.27953004837036	-1.76808998730849e-06	7.29826979295467e-06	1.22642995847855e-05	0.936984002590179	-999.0	-999.0	-9.52444042923162e-06	1.15217999336892e-05	2.78790994343581e-05	0.901268005371094	-999.0	-999.0	114720	-0.125719994306564	-0.113649003207684	16	8	8	8	8	8	8	1.01845998869976e-05	6.45187981263007e-07	0.308281004428864	21.380199432373	0.0687810033559799	6.69680002829409e-06	1.69485997503216e-06	1.0584499835968	21.8353996276855	0.274783998727798	6.33960007689893e-06	2.91953006126278e-06	5.07843017578125	21.8948993682861	0.500006973743439	5.26216990692774e-06	7.46496016290621e-06	9.11267034098273e-06	0.946640014648438	22.0972003936768	1.54024004936218	-9.60307033892605e-07	1.17714998850715e-05	1.9690200133482e-05	0.98087602853775	-999.0	-999.0	-1.54449007823132e-05	1.88798003364354e-05	2.58110994764138e-05	0.977401971817017	-999.0	-999.0	114720	-0.287185996770859	-0.523117005825043	23	6	6	6	6	6	6	1.77762995008379e-05	2.16220996662742e-06	0.491183996200562	20.7754993438721	0.132063001394272	7.29061002857634e-06	7.67850997362984e-06	2.04662990570068	21.743200302124	1.1434999704361	8.09812991064973e-06	6.76764011586783e-06	6.99106979370117	21.6291007995605	0.90735399723053	2.71738002766142e-07	1.81544000952272e-05	1.77294004970463e-05	0.974824011325836	25.3146991729736	72.5362014770508	-2.27357995754573e-05	2.84660000033909e-05	5.92994001635816e-05	1.00258994102478	-999.0	-999.0	-5.46818992006592e-05	4.55734989373013e-05	9.56232979660854e-05	1.00075995922089	-999.0	-999.0	114720	0.361384004354477	-0.239363998174667
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
QSO	0.13154528	0.048649073	0.8198056	-999.0	-999.0	215802945600502567	4014152000004343	361670606105105	0.589556147893477	76	4014152	3	11	7	7	5	7	7	7	6.3835700530035e-06	9.08780975805712e-07	0.771224975585938	21.8873996734619	0.154568001627922	6.07162019150564e-06	3.39877010446799e-06	4.93940019607544	21.9417991638184	0.607773005962372	1.07454998214962e-05	1.89864999811107e-06	3.904629945755	21.32200050354	0.191841006278992	6.6823899942392e-06	3.370100102984e-06	5.41632016393123e-06	0.97071897983551	21.8376998901367	0.547564029693604	7.61141018301714e-06	5.25808991369559e-06	1.3404000128503e-05	0.980018019676209	21.6963996887207	0.750043988227844	1.02794001577422e-05	8.42367990117054e-06	2.38649990933482e-05	0.960997998714447	21.3701000213623	0.889730989933014	16416	0.0200907997786999	-0.4706209897995	16	12	12	11	12	12	12	9.44018029258586e-06	1.00770000699413e-06	1.12676000595093	21.4626007080078	0.115897998213768	8.45168960950105e-06	1.80171002739371e-06	2.41034007072449	21.5827007293701	0.231453999876976	8.16857027530205e-06	3.74922001356026e-06	14.0581998825073	21.6196994781494	0.498331993818283	8.64694993651938e-06	5.17867010785267e-06	7.02287979947869e-06	0.974824011325836	21.5578994750977	0.650249004364014	5.96194013269269e-06	8.10993969935225e-06	1.55498000822263e-05	0.989114999771118	21.9615993499756	1.47690999507904	8.86395991983591e-06	1.2925500413985e-05	3.01395994029008e-05	0.952482998371124	21.5310001373291	1.58323001861572	114720	0.0723079964518547	-0.214270994067192	21	12	12	11	12	12	12	9.8328400781611e-06	8.45784995817667e-07	0.813528001308441	21.418399810791	0.0933910012245178	1.16440996862366e-05	1.76272999397042e-06	1.30164003372192	21.2348003387451	0.164361998438835	1.22946003102697e-05	3.37197002409084e-06	10.3430004119873	21.1758003234863	0.297778010368347	1.19093001558213e-05	6.94272011969588e-06	8.88861995917978e-06	0.96450799703598	21.2103996276855	0.632947981357574	1.06959996628575e-05	1.08023996290285e-05	2.9420800274238e-05	0.969403028488159	21.3269996643066	1.09653997421265	1.17311001304188e-05	1.71875999512849e-05	5.68737013963982e-05	0.948948979377747	21.2266998291016	1.59073996543884	114720	0.0825961008667946	0.19777500629425	11	1	1	1	0	0	0	1.55571997311199e-05	5.78718982069404e-06	-999.0	20.9202003479004	0.403889000415802	0.0	0.0	-999.0	-999.0	-999.0	5.97892994846916e-06	5.78718982069404e-06	-999.0	21.9584999084473	1.05092000961304	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	16416	-999.0	-999.0	17	5	5	5	5	5	5	1.89227994269459e-05	4.28774001193233e-06	0.39309498667717	20.7075996398926	0.246018007397652	5.02793009218294e-05	9.21065020520473e-06	0.478060007095337	19.6466007232666	0.198896005749702	4.25723992520943e-05	9.08127003640402e-06	2.51114988327026	19.8271999359131	0.231601998209953	4.74555017717648e-05	2.5511499188724e-05	1.42212002174347e-05	0.974824011325836	19.709400177002	0.583678007125854	5.15071005793288e-05	3.99558994104154e-05	5.12396982230712e-05	1.00258994102478	19.6203994750977	0.842244982719421	6.04583983658813e-05	6.40326979919337e-05	0.000174567001522519	0.99857097864151	19.4463996887207	1.14991998672485	16416	0.286370992660522	0.694639980792999
STAR	7.238483e-05	0.9999058	2.17957e-05	-999.0	-999.0	215811127390347511	4051687000001124	361662016136093	0.82680821961426	91	4051687	3	13	9	8	7	7	7	7	0.00016390900418628	1.84955001714116e-06	2.48726010322571	18.3635997772217	0.0122509999200702	0.000156263005919755	1.49028994655964e-06	1.43408000469208	18.4153995513916	0.0103550003841519	0.000164633005624637	1.71788997249678e-06	3.23148989677429	18.3588008880615	0.0113289998844266	0.000475439999718219	9.18216028367169e-06	0.00025128701236099	0.974824011325836	17.2073001861572	0.0209689997136593	0.000491953978780657	1.39173998832121e-05	0.000245712988544256	1.00258994102478	17.1702995300293	0.0307160001248121	0.000480307004181668	2.18654004129348e-05	0.00022173099569045	0.99211198091507	17.1963005065918	0.0494269989430904	115000	-0.0171202998608351	0.0308749992400408	21	11	12	11	7	7	7	0.000346271001035348	1.94811991605093e-06	3.14560008049011	17.5515003204346	0.00610799994319677	0.000319000013405457	2.73022010333079e-06	5.3051700592041	17.6406002044678	0.00929199997335672	0.000341487000696361	2.42511009673763e-06	5.69199991226196	17.5666007995605	0.00771000003442168	0.000341264996677637	7.16877002560068e-06	6.92497997079045e-05	0.974824011325836	17.5673007965088	0.0228070002049208	0.000372674985555932	1.09276998045971e-05	7.41655021556653e-05	0.985430002212524	17.471700668335	0.031835999339819	0.000370144000044093	1.64168995979708e-05	0.000109920001705177	0.924646973609924	17.4790992736816	0.0481549985706806	115000	0.005839420016855	0.0145984999835491	27	11	11	10	9	9	9	0.000480304006487131	1.91851995623438e-06	1.35175001621246	17.1963005065918	0.00433700019493699	0.000441409996710718	3.82177995561506e-06	7.24265003204346	17.2880001068115	0.00939999986439943	0.000480232993140817	5.29437011209666e-06	15.8977003097534	17.1963996887207	0.0119700003415346	0.000158238995936699	7.11469010639121e-06	0.000143821001984179	0.697040021419525	18.4018001556396	0.0488159991800785	0.000161417992785573	1.0920600288955e-05	0.000155817993800156	0.726608991622925	18.380199432373	0.0734549984335899	0.000151708998600952	1.76680005097296e-05	0.000166117999469861	0.745446026325226	18.4475002288818	0.12644399702549	115000	0.00475687021389604	0.00669485982507467	12	9	9	7	6	6	6	0.000571387994568795	2.38768006965984e-06	1.76490998268127	17.0076999664307	0.00453699985519052	0.000531048979610205	5.0977600949409e-06	2.69152998924255	17.0872001647949	0.0104219997301698	0.000568975985515863	4.49610979558202e-06	4.05645990371704	17.012300491333	0.00858000013977289	0.000267344003077596	8.0743102444103e-06	0.000260939996223897	0.82709002494812	17.8323993682861	0.0327909998595715	0.00027947299531661	1.19912001537159e-05	0.00028328801272437	0.821016013622284	17.784200668335	0.0465850010514259	0.000286216003587469	1.84942000487354e-05	0.000306222005747259	0.797819018363953	17.75830078125	0.0701560005545616	115000	-0.00171305995900184	-0.0218414999544621	18	5	5	4	5	5	5	0.000640467973425984	3.43222995979886e-06	0.737075984477997	16.8838005065918	0.00581800006330013	0.000547080999240279	9.0059002104681e-06	2.83067011833191	17.0548992156982	0.0178730003535748	0.000618058023974299	1.414360031049e-05	4.69218015670776	16.9225006103516	0.0248460005968809	0.000917459023185074	1.83488991751801e-05	0.000786132994107902	0.769690990447998	16.4936008453369	0.021714000031352	0.000945253006648272	2.81567008642014e-05	0.000821072026155889	0.793201982975006	16.4612007141113	0.0323409996926785	0.000940750003792346	4.43669014202897e-05	0.000819849024992436	0.765676975250244	16.4664001464844	0.05120500177145	115000	-0.00836171954870224	-0.0213687997311354
STAR	0.002416611	0.9943745	0.0032088761	-999.0	-999.0	215811516798300902	4051687000001929	361662016142264	0.554106026472623	91	4051687	3	13	7	8	8	4	4	4	0.00293629011139274	5.24629012943478e-06	2.82296991348267	15.2306003570557	0.00193999998737127	0.00285206991247833	9.53196013142588e-06	11.7985000610352	15.2622003555298	0.00362899992614985	0.00294223008677363	6.49583989797975e-06	4.17322015762329	15.2284002304077	0.0023970000911504	0.0105865001678467	1.5658299162169e-05	0.00593284005299211	0.974824011325836	13.8381996154785	0.00160600000526756	0.0111352996900678	1.97386998479487e-05	0.00614470010623336	1.00258994102478	13.7833003997803	0.00192499998956919	0.0113682998344302	2.74776994046988e-05	0.00622011022642255	1.00093996524811	13.7608003616333	0.00262399995699525	115000	-0.0109740998595953	0.0281805004924536	21	9	9	9	10	10	10	0.00627719005569816	1.64205994224176e-05	20.9612007141113	14.4056997299194	0.00283999997191131	0.00606695003807545	2.05320993700298e-05	52.3573989868164	14.4426002502441	0.0036740000359714	0.0063539301045239	1.71354004123714e-05	25.3841991424561	14.3924999237061	0.0029279999434948	0.00462277978658676	7.95551022747532e-06	0.00340139004401863	0.935920000076294	14.7377996444702	0.00186800002120435	0.00489648012444377	9.72831003309693e-06	0.00353271001949906	0.964704990386963	14.6753997802734	0.0021569998934865	0.00502035999670625	1.31400001919246e-05	0.00360525003634393	0.973052978515625	14.6482000350952	0.0028419999871403	115000	-0.016619399189949	0.0232295002788305	27	8	12	10	11	11	11	0.00902056973427534	2.00188005692326e-05	23.8878002166748	14.0120000839233	0.00240999995730817	0.00859041046351194	2.12155991903273e-05	71.7469024658203	14.0649995803833	0.00268099992536008	0.00906398985534906	1.74628003151156e-05	21.5533008575439	14.006799697876	0.00209200009703636	0.00499291997402906	1.08906997411395e-05	0.00237704999744892	0.974824011325836	14.6541996002197	0.00236800010316074	0.00534967007115483	1.42709995998302e-05	0.002452929969877	0.996459007263184	14.5791997909546	0.00289599993266165	0.00551712978631258	2.02751998585882e-05	0.00246808002702892	0.968106985092163	14.5458002090454	0.00399000011384487	115000	0.00262166000902653	-0.0167130995541811	12	2	2	2	1	1	1	0.0108497999608517	0.000176037996425293	184.255996704102	13.8114995956421	0.0176160000264645	0.0101589998230338	8.50395008455962e-05	53.4873008728027	13.8829002380371	0.00908899959176779	0.0108041996136308	2.00751001102617e-05	2.3910698890686	13.8161001205444	0.00201699999161065	0.00386289996095002	1.50424002640648e-05	-999.0	0.963771998882294	14.9328002929688	0.00422800006344914	0.00419188011437655	2.14941992453532e-05	-999.0	0.988731026649475	14.8439998626709	0.0055669997818768	0.00447783013805747	3.29583017446566e-05	-999.0	0.995552003383636	14.7723999023438	0.00799100007861853	115000	-0.0599368996918201	0.0977917984127998	18	9	11	11	7	7	7	0.0121812997385859	4.49330000265036e-05	18.9601993560791	13.6857995986938	0.00400499999523163	0.011455399915576	2.31220001296606e-05	4.17178010940552	13.7524995803833	0.00219099991954863	0.0120951998978853	2.84405996353598e-05	5.47292995452881	13.6934995651245	0.00255300011485815	0.0349926985800266	5.84699009777978e-05	0.0330867990851402	0.904088020324707	12.5401000976563	0.00181399995926768	0.0375524014234543	7.28414015611634e-05	0.0346142984926701	0.950052976608276	12.4635000228882	0.00210599997080863	0.0386836007237434	9.98139003058895e-05	0.0351576991379261	0.965413987636566	12.4312000274658	0.00280099990777671	115000	0.00154083000961691	0.028575399890542
QSO	0.0017456317	0.059080012	0.9391744	-999.0	-999.0	215811728488342412	4051687000002251	361662016144941	0.417286351711756	91	4051687	3	13	8	8	7	7	7	7	1.39362000481924e-05	6.02067018462549e-07	0.449014991521835	21.0396995544434	0.0469059981405735	1.54856006702175e-05	1.07812002170249e-06	2.00268006324768	20.925199508667	0.0755899995565414	1.38403001983534e-05	1.10649000362173e-06	1.59735000133514	21.0471992492676	0.0868009999394417	3.54955009242985e-05	1.10132996269385e-05	1.38492996484274e-05	0.974824011325836	20.0245990753174	0.33687698841095	4.8060501285363e-05	1.70582006830955e-05	3.01427007798338e-05	1.00258994102478	19.6956005096436	0.3853619992733	2.58416002907325e-05	2.69941992883105e-05	7.87734024925157e-05	0.998633027076721	20.3693008422852	1.13416004180908	115000	0.0244954992085695	0.151982992887497	21	10	10	9	4	4	4	1.94749991351273e-05	1.65723997724854e-06	1.97143995761871	20.676399230957	0.0923919975757599	1.45047997648362e-05	1.65059998380457e-06	2.42406010627747	20.9962997436523	0.12355399876833	1.82973999471869e-05	2.63425999946776e-06	7.17428016662598	20.7441005706787	0.156312003731728	3.12181982735638e-05	7.11102984496392e-06	3.30084985762369e-05	0.974824011325836	20.1639995574951	0.247315004467964	4.91216997033916e-05	1.09675002022414e-05	5.51291013834998e-05	0.975100994110107	19.6718997955322	0.242413997650146	9.95899972622283e-05	1.71872998180334e-05	0.000119465999887325	0.933194994926453	18.9044990539551	0.187377005815506	115000	-0.277254998683929	0.298283994197845	27	17	17	16	9	9	9	2.57368992606644e-05	8.11743007034238e-07	0.941026985645294	20.3736991882324	0.0342440009117126	2.40603003476281e-05	1.86726003903459e-06	1.70977997779846	20.4468002319336	0.0842610001564026	2.80276999546913e-05	2.40380995819578e-06	7.37856006622314	20.281099319458	0.0931190028786659	1.7004000255838e-05	5.87904014537344e-06	1.5891400835244e-05	0.866509974002838	20.8236999511719	0.375387996435165	1.92519000847824e-05	9.75085004029097e-06	3.46591987181455e-05	0.8923220038414	20.6888999938965	0.549911022186279	2.98357008432504e-05	2.09085992537439e-05	6.77209973218851e-05	0.898473024368286	20.2131996154785	0.760873019695282	115000	0.0810011997818947	-0.112810999155045	12	8	8	7	5	5	5	2.84801008092472e-05	2.11840006159036e-06	2.00382995605469	20.2637004852295	0.0807589963078499	2.32719994528452e-05	1.96103997041064e-06	1.34274995326996	20.4829998016357	0.0914909988641739	1.22120000014547e-05	6.51335994916735e-06	12.7267999649048	21.1830997467041	0.579086005687714	1.44565001392039e-05	7.89177011029096e-06	2.66497991106007e-05	0.974824011325836	20.9999008178711	0.592700004577637	2.25552994379541e-05	1.22682004075614e-05	4.95230997330509e-05	0.984849989414215	20.5168991088867	0.590550005435944	3.59797013516072e-05	1.93410996871535e-05	0.00011380499927327	0.956161022186279	20.0098991394043	0.583644986152649	115000	-0.169009998440742	0.15370200574398	18	8	8	8	7	7	7	4.41968004452065e-05	7.42100019124337e-06	3.42902994155884	19.786600112915	0.182303994894028	5.98178994550835e-05	1.18807001854293e-05	6.60586023330688	19.4580001831055	0.215643003582954	0.000111323002784047	2.04906991712051e-05	36.1575012207031	18.7835998535156	0.199845999479294	5.38837011845317e-05	1.87665991688846e-05	4.80554990645032e-05	0.835563004016876	19.5713996887207	0.378138989210129	7.24807978258468e-05	2.97114002023591e-05	9.76668015937321e-05	0.872829020023346	19.2495002746582	0.44506499171257	6.9341498601716e-05	4.8352998419432e-05	0.000142680000863038	0.890964984893799	19.2975997924805	0.757102012634277	16892216	-1.49083995819092	0.261462986469269
GALAXY	0.8302878	0.019351477	0.15036066	0.288313284516335	0.164239302982127	215833405070605911	4014152000005063	361670606105829	0.271554662167611	76	4014152	3	11	1	1	1	1	1	1	3.86058991352911e-06	2.46629997491254e-06	-999.0	22.4333992004395	0.69361400604248	2.6372299544164e-06	3.95944016418071e-06	-999.0	22.8472003936768	1.63007998466492	-2.68909002443252e-06	2.46629997491254e-06	-999.0	-999.0	-999.0	9.26298014292115e-07	2.82773999060737e-06	-999.0	0.974824011325836	23.9832000732422	3.31446003913879	3.92976005514356e-07	4.44087982032215e-06	-999.0	1.00258994102478	24.9141998291016	12.2694997787476	-1.55807003920927e-06	7.19116997061064e-06	-999.0	1.00093996524811	-999.0	-999.0	114720	-0.294117987155914	-1.25821995735168	16	2	2	2	2	2	2	4.61402987639303e-06	3.30875991494395e-06	1.95288002490997	22.2399005889893	0.778590023517609	9.33496994548477e-06	7.77034983912017e-06	3.2925500869751	21.4748001098633	0.903757989406586	1.18035995910759e-05	1.18197999654512e-06	0.248934999108315	21.2199993133545	0.108722999691963	6.88447016727878e-06	5.46557021152694e-06	3.27691009260889e-06	0.974824011325836	21.8054008483887	0.861965000629425	1.25652004498988e-05	8.5742703959113e-06	5.49826017959276e-06	1.00258994102478	21.152099609375	0.740884006023407	2.55922004726017e-05	1.37373999677948e-05	6.24594986220472e-06	1.00093996524811	20.3798007965088	0.582800984382629	16416	0.291177004575729	-0.182475000619888	21	4	3	3	4	4	4	9.33622959564673e-06	3.27619000017876e-06	2.37278008460999	21.4745998382568	0.380997002124786	6.86033990859869e-06	2.20622005144833e-05	7.88825988769531	21.8092002868652	3.49163007736206	1.06032002804568e-05	1.08144995465409e-05	18.917200088501	21.3365001678467	1.10738003253937	1.61633997777244e-05	7.20135994924931e-06	1.53573000716278e-05	0.974824011325836	20.8787002563477	0.483734011650085	2.81426000583451e-05	1.13184996735072e-05	3.6003599234391e-05	1.00213003158569	20.2766990661621	0.436666011810303	5.22709015058354e-05	1.81032992259134e-05	5.4209798690863e-05	0.996630012989044	19.6044006347656	0.376029014587402	16416	0.244028002023697	0.126539006829262	11	2	2	2	2	2	2	3.15475995193992e-06	6.39353004316945e-07	0.0518474988639355	22.6527004241943	0.220038995146751	1.3954099813418e-05	1.99983005586546e-05	3.64861989021301	21.0382995605469	1.55603003501892	8.73076987772947e-06	9.28065037442138e-06	11.773099899292	21.547399520874	1.15411996841431	5.85754014537088e-06	7.06409991835244e-06	1.99338992388221e-05	0.974824011325836	21.9808006286621	1.30938005447388	6.17665000390843e-06	1.10824003058951e-05	1.97358003788395e-05	1.00258994102478	21.9232006072998	1.94807994365692	-1.24956000036036e-06	1.77454003278399e-05	5.09126984979957e-05	1.00093996524811	-999.0	-999.0	16416	7.11289978027344	-42.7580986022949	17	2	2	2	2	2	2	1.5274299585144e-05	2.50867997237947e-05	4.8086199760437	20.9402008056641	1.78323996067047	3.2160500268219e-05	4.12349982070737e-06	0.0405201017856598	20.131799697876	0.139209002256393	3.91456014767755e-05	2.7916799808736e-05	5.98519992828369	19.918399810791	0.774294972419739	4.94275991513859e-05	2.56069997703889e-05	3.35568984155543e-05	0.944983005523682	19.6651000976563	0.5624880194664	4.86452008772176e-05	3.74521005142014e-05	1.26211000406329e-06	0.902332007884979	19.6825008392334	0.835913002490997	-4.17072988057043e-06	5.64184010727331e-05	9.16741992114112e-05	0.84747701883316	-999.0	-999.0	16416	-0.267080992460251	-0.198758006095886
UNSURE	0.28027123	0.22520801	0.49452075	-999.0	-999.0	215840436920662798	4047197000001602	361657721171416	0.359974941574014	65	4047197	3	9	1	1	1	0	0	0	7.27905990061117e-06	4.03177000407595e-06	-999.0	21.7448997497559	0.601373970508575	-1.00794995887554e-05	7.39398001314839e-06	-999.0	-999.0	-999.0	-8.76314970810199e-06	4.03177000407595e-06	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	16416	-999.0	-999.0	16	9	9	9	7	7	7	9.22112030821154e-06	6.94690015734523e-07	0.652456998825073	21.4881000518799	0.0817959979176521	8.40578013594495e-06	1.15601994821191e-06	0.9269899725914	21.5886001586914	0.149316996335983	6.09694006925565e-06	1.30983994495182e-06	1.9789799451828	21.9372997283936	0.233255997300148	6.53719007459586e-06	5.32752983417595e-06	7.70636961533455e-06	0.835563004016876	21.8616008758545	0.884829998016357	1.2299399713811e-05	8.32944988360396e-06	2.23875995288836e-05	0.854611992835999	21.1753997802734	0.735288977622986	8.09972971183015e-06	1.31587003124878e-05	2.96220005111536e-05	0.831354975700378	21.6289005279541	1.76388001441956	114736	0.0121585000306368	0.0743812993168831	12	7	7	6	5	5	5	1.26739996630931e-05	8.6284097733369e-07	0.658354997634888	21.1427993774414	0.0739160031080246	1.02277999758371e-05	4.5613301153935e-06	4.58004999160767	21.3756008148193	0.484209001064301	8.52854009281145e-06	3.63840990758035e-06	8.33868980407715	21.5729007720947	0.463192999362946	1.90519995157956e-06	4.51689993496984e-06	6.70824010740034e-06	0.779859006404877	23.2001991271973	2.57407999038696	9.53884978116548e-07	7.01578983353102e-06	8.93257038114825e-06	0.781929016113281	23.9512996673584	7.98555994033813	4.23270012106514e-06	1.10091996248229e-05	2.26501997531159e-05	0.726213991641998	22.3334999084473	2.82398009300232	115000	-0.618142008781433	-0.89148998260498	11	6	6	6	3	3	3	9.84714006335707e-06	2.10852999771305e-06	1.62476003170013	21.4167995452881	0.232483997941017	6.46099988443893e-06	3.36594007421809e-06	1.9190000295639	21.8743000030518	0.565627992153168	8.01291025709361e-06	1.82755002242629e-06	2.03625011444092	21.6406002044678	0.247630000114441	1.48722001540591e-05	7.68868994782679e-06	2.43393005803227e-05	0.974824011325836	20.9691009521484	0.561307013034821	3.84023005608469e-05	1.19523001558264e-05	3.73770017176867e-05	1.00258994102478	19.9391994476318	0.337924003601074	3.19565988320392e-05	1.89742004295113e-05	5.90704985370394e-05	0.999026000499725	20.1387004852295	0.644653975963593	16416	0.128151997923851	-0.687596023082733	17	5	5	5	4	4	4	2.5744100639713e-05	1.01299001471489e-05	2.29470992088318	20.3733997344971	0.427219986915588	6.99551019351929e-05	2.19171997741796e-05	3.82873010635376	19.2880001068115	0.340164989233017	8.22540969238617e-05	3.01973996101879e-05	20.0414009094238	19.1121997833252	0.398597985506058	9.52573009271873e-06	1.56053993123351e-05	3.42821003869176e-05	0.73167097568512	21.4528007507324	1.77869999408722	-3.23164999826986e-06	2.60043998423498e-05	2.42458008870017e-05	0.7932950258255	-999.0	-999.0	-1.29725003716885e-05	4.51300984423142e-05	4.25590005761478e-05	0.85206001996994	-999.0	-999.0	16416	0.0738359987735748	-0.142980992794037
QSO	0.09862803	0.039303366	0.86206853	-999.0	-999.0	215881712479949329	4051687000002004	361662016142735	0.541622547005883	165	4051687	3	22	6	5	5	10	10	10	8.11224981589476e-06	9.21509979434632e-07	0.877155005931854	21.6271991729736	0.123333998024464	1.11348999780603e-05	2.3293100639421e-06	1.45432996749878	21.2833003997803	0.227125003933907	1.18485995699302e-05	2.12994996218185e-06	3.2334098815918	21.2159004211426	0.195175006985664	9.60530996962916e-06	5.60523994863615e-06	7.67563960835105e-06	0.798205971717834	21.4437999725342	0.633588016033173	1.49718998727622e-05	8.76851026987424e-06	1.34981000883272e-05	0.82591301202774	20.9619007110596	0.635877013206482	2.47737007157411e-05	1.38573996082414e-05	2.87519997073105e-05	0.829865992069244	20.4151000976563	0.607315003871918	16416	-0.0768020004034042	0.261451005935669	35	9	9	8	11	11	11	1.05848002931452e-05	1.176259956992e-06	1.10564005374908	21.338399887085	0.120655998587608	8.56712995300768e-06	1.2363200312393e-06	1.32876002788544	21.568000793457	0.156681999564171	1.02277999758371e-05	2.35526999858848e-06	5.46781015396118	21.3756008148193	0.250023007392883	7.76988963480107e-06	4.23976007368765e-06	5.88960983805009e-06	0.88620400428772	21.673999786377	0.592449009418488	4.15410977439024e-06	6.67229005557601e-06	1.35137997858692e-05	0.916655004024506	22.3539009094238	1.7438999414444	-9.03952968656085e-06	1.08429003375932e-05	2.26635002036346e-05	0.93471497297287	-999.0	-999.0	115000	-0.111308999359608	0.195879995822906	51	16	16	16	27	27	27	1.53561995830387e-05	1.08505003026949e-06	0.914686977863312	20.9344005584717	0.0767169967293739	1.41416003316408e-05	2.12861004911247e-06	1.49040997028351	21.0237998962402	0.163426995277405	1.54857007146347e-05	2.39115001932078e-06	6.42195987701416	20.925199508667	0.167648002505302	1.21817001854652e-05	8.57854956848314e-06	1.35568998302915e-05	0.974824011325836	21.1858005523682	0.764591991901398	9.58450982579961e-06	1.34641004478908e-05	3.13232012558728e-05	1.00258994102478	21.4461002349854	1.52522003650665	1.4647799616796e-05	2.16079006349901e-05	5.98750011704396e-05	1.00093996524811	20.9855995178223	1.60163998603821	115000	-0.0814921036362648	0.0316704995930195	22	5	5	4	8	8	8	1.465589957661e-05	2.9169900699344e-06	1.47581994533539	20.9850006103516	0.216095998883247	1.88320991583169e-05	8.71151041792473e-06	4.05038022994995	20.7127990722656	0.502249002456665	1.43849001688068e-05	3.43710007655318e-06	2.99870991706848	21.0053005218506	0.25942400097847	1.41025002449169e-05	2.18333007069305e-05	1.54823992488673e-05	0.892207026481628	21.0268001556396	1.68092000484467	1.06140996649629e-05	3.60369995178189e-05	2.4984899937408e-05	0.901408016681671	21.3353996276855	3.68630003929138	-4.35647007179796e-06	5.31471014255658e-05	4.49266990472097e-05	0.822543978691101	-999.0	-999.0	16416	-0.5521320104599	-0.0992475971579552	35	8	8	8	14	14	14	2.46020008489722e-05	5.57478006157908e-06	0.877574980258942	20.4225997924805	0.246025994420052	1.50349997056765e-05	1.17564995889552e-05	3.08294010162354	20.9573001861572	0.848980009555817	1.51716003529145e-05	9.98128962237388e-06	3.20192003250122	20.9475002288818	0.714295983314514	4.83426010760013e-05	3.07996997435112e-05	4.74865992146079e-05	0.971350014209747	19.6891994476318	0.691736996173859	5.90372983424459e-05	4.795740096597e-05	6.91995010129176e-05	0.977707982063293	19.4722003936768	0.881968975067139	0.00011195799743291	7.64095966587774e-05	0.000176410001586191	0.945261001586914	18.7773990631104	0.74099999666214	16416	-0.045101098716259	-0.431942999362946
UNSURE	0.6745695	0.14276634	0.18266417	-999.0	-999.0	215900249688286907	4047197000005516	361657721201206	0.147580708514509	65	4047197	3	9	3	3	2	3	3	3	1.51493004523218e-05	4.95319000037853e-06	6.4297399520874	20.9491004943848	0.354988992214203	2.943649997178e-05	7.76458000473212e-06	5.68076992034912	20.2278003692627	0.286388009786606	3.91899993701372e-05	3.86205010727281e-06	3.17767000198364	19.9171009063721	0.10699599981308	2.20435003939201e-06	5.44901013199706e-06	6.12518988418742e-06	0.649882972240448	23.0419006347656	2.68387007713318	1.01012001323397e-05	8.63405966811115e-06	1.50640998981544e-05	0.668394029140472	21.3890991210938	0.928043007850647	1.60878994392988e-06	1.38121004056302e-05	7.22538015907048e-06	0.663941025733948	23.3838005065918	9.32147026062012	114720	-0.147054001688957	0.29660901427269	16	2	2	2	1	1	1	5.58750980417244e-06	1.45291005537729e-06	0.309208005666733	22.0319995880127	0.282321989536285	6.91350987835904e-06	1.18014997951832e-06	0.0584188997745514	21.8008003234863	0.185337007045746	8.38016967463773e-06	2.82413998320408e-06	1.16851997375488	21.5918998718262	0.365895003080368	1.56078003783477e-05	3.87806994694984e-06	-999.0	0.974824011325836	20.9167003631592	0.269773006439209	3.40842016157694e-05	6.07174979450065e-06	-999.0	1.00258994102478	20.0687007904053	0.193413004279137	5.20411012985278e-05	9.72794987319503e-06	-999.0	1.00093996524811	19.6091995239258	0.202954992651939	16416	-0.223605006933212	0.0433560982346535	12	2	2	2	0	0	0	1.58548004947079e-07	9.59414023782301e-07	0.0511070005595684	25.8997001647949	6.57004976272583	-9.09091977518983e-06	1.54137997014914e-05	2.7563099861145	-999.0	-999.0	-2.61436002801929e-06	1.6641199181322e-05	17.1615009307861	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	16416	-999.0	-999.0	11	2	2	2	1	1	1	8.55971029523062e-06	2.27674991037929e-06	0.395857989788055	21.5688991546631	0.288789004087448	3.28774003719445e-05	1.16049995995127e-05	0.924902975559235	20.1077995300293	0.38324099779129	1.20578997666598e-05	1.62866999744438e-05	20.4640998840332	21.1968994140625	1.46651005744934	2.89139006781625e-06	8.40868960949592e-06	-999.0	0.974824011325836	22.7472991943359	3.15752005577087	-1.16012997750659e-05	1.31780998344766e-05	-999.0	1.00258994102478	-999.0	-999.0	-5.71046002733056e-05	2.11340993701015e-05	-999.0	1.00021994113922	-999.0	-999.0	16416	-1.68329000473022	2.18952989578247	17	3	3	3	0	0	0	4.66938990939525e-06	2.45165006163006e-06	0.134582996368408	22.226900100708	0.57006299495697	2.50342000072123e-05	3.21715997415595e-05	4.1339898109436	20.4036998748779	1.39528000354767	-1.12397998464076e-07	3.05950015899725e-05	12.8643999099731	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	16416	-999.0	-999.0
UNSURE	0.11359814	0.28135583	0.60504603	-999.0	-999.0	215931594416049445	4051687000001756	361662016140877	0.647658265798966	91	4051687	3	13	3	2	2	0	0	0	7.60400007493445e-06	3.16770001518307e-06	7.98159980773926	21.6975002288818	0.452298998832703	2.72066008619731e-05	5.11840016770293e-06	3.93690991401672	20.3134002685547	0.204260006546974	2.55932991422014e-05	2.38907000493782e-06	2.95379996299744	20.3798007965088	0.101351000368595	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	16892216	-999.0	-999.0	21	4	3	3	3	3	3	2.94491997010482e-06	2.17259002965875e-06	2.5904700756073	22.7273998260498	0.800993978977203	7.79929996497231e-06	4.82167979498627e-06	4.225830078125	21.669900894165	0.671223998069763	-4.16866987507092e-06	4.6334798753378e-06	4.69747018814087	-999.0	-999.0	3.5056700653513e-07	5.18850993103115e-06	1.27967996377265e-05	0.66093498468399	25.0380992889404	16.069299697876	-1.61246998686693e-06	8.39663971419213e-06	1.45216999953846e-05	0.726917028427124	-999.0	-999.0	-2.24283994612051e-05	1.38397999762674e-05	5.11721991642844e-05	0.768527984619141	-999.0	-999.0	16416	-0.0307941995561123	-0.871084988117218	27	7	6	6	3	3	3	6.37801986158593e-06	1.64607001806871e-06	1.14728999137878	21.8883991241455	0.280211001634598	2.26896991080139e-05	6.4675500652811e-06	3.29941010475159	20.5104999542236	0.309480994939804	3.04565000988077e-05	6.0811898947577e-06	7.58496999740601	20.1909008026123	0.216786995530128	1.15780994747183e-06	8.24512972030789e-06	2.54021997534437e-05	0.840721011161804	23.7409992218018	7.73188018798828	1.02932999652694e-05	1.29004001792055e-05	4.66603014501743e-05	0.854744017124176	21.3687000274658	1.36073005199432	1.30265998450341e-05	2.06509994313819e-05	0.000110943001345731	0.850831985473633	21.113000869751	1.72121000289917	16416	-0.855605006217957	1.44457995891571	12	1	1	1	0	0	0	5.2894902182743e-06	5.90881018069922e-06	-999.0	22.0914993286133	1.21285998821259	-0.000105844002973754	2.77685994660715e-05	-999.0	-999.0	-999.0	-1.27970997709781e-05	5.90881018069922e-06	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	16416	-999.0	-999.0	18	3	3	2	0	0	0	1.16121000246494e-05	2.02380997507134e-05	3.0813000202179	21.2378005981445	1.89226996898651	6.92038011038676e-05	7.9132296377793e-05	18.4437999725342	19.2996997833252	1.24150002002716	8.61053995322436e-05	3.43540996254887e-05	9.0955696105957	19.0625	0.433183997869492	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	16416	-999.0	-999.0
STAR	0.013104182	0.97511476	0.011781078	-999.0	-999.0	215972212048211215	3937891000001521	361666311107137	0.882625562029542	165	3937891	3	22	6	6	5	0	0	0	4.27990016760305e-05	1.21689004117798e-06	1.5507800579071	19.8215007781982	0.0308699999004602	3.89335000363644e-05	2.00596991817292e-06	2.73924994468689	19.9242992401123	0.0559399984776974	4.11736982641742e-05	1.92119000530511e-06	3.22670006752014	19.8635005950928	0.0506610013544559	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	-999.0	115000	-999.0	-999.0	35	10	10	8	3	3	3	7.49929968151264e-05	1.31679996684397e-06	1.48949003219604	19.2124996185303	0.01906399987638	6.75120027153753e-05	2.51506003223767e-06	2.61616992950439	19.3265991210938	0.0404479987919331	7.04733029124327e-05	3.04665991279762e-06	6.9282398223877	19.2800006866455	0.0469379983842373	4.32755005022045e-05	7.7781096479157e-06	3.03468004858587e-05	0.881246984004974	19.8094997406006	0.19514499604702	3.37523015332408e-05	1.19982996693579e-05	4.49955987278372e-05	0.882465004920959	20.0792999267578	0.385957986116409	2.64978007180616e-05	1.89988004422048e-05	8.04638984845951e-05	0.863170981407166	20.3419990539551	0.778469979763031	115000	-0.0277392007410526	-0.0658807009458542	51	17	17	16	8	8	8	9.40199970500544e-05	1.29788998037839e-06	1.17748999595642	18.9669990539551	0.01498799957335	8.72375967446715e-05	2.38357006310252e-06	2.80322003364563	19.0482997894287	0.0296650007367134	9.61062032729387e-05	1.99773990061658e-06	2.36469006538391	18.9431991577148	0.0225690007209778	8.44411988509819e-05	6.94766004016856e-06	5.12459992023651e-05	0.949955999851227	19.0837001800537	0.0893319994211197	8.89518996700644e-05	1.06593997770688e-05	5.2115399739705e-05	0.963319003582001	19.0272006988525	0.130107000470161	7.84239018685184e-05	1.67718008015072e-05	5.63971989322454e-05	0.949234008789063	19.1639003753662	0.232197001576424	115000	0.00799410045146942	-0.049440398812294	22	7	7	6	2	2	2	0.000113617003080435	9.48377021359192e-07	0.395862996578217	18.7614994049072	0.00906299985945225	0.000111988003482111	1.12514999273117e-06	0.167748004198074	18.777099609375	0.0109080001711845	0.000121343997307122	1.549949956825e-06	0.506164014339447	18.6900005340576	0.0138680003583431	7.03939003869891e-05	1.20422000691178e-05	1.76415996975265e-05	0.974824011325836	19.2812004089355	0.185736000537872	8.75601035659201e-05	1.88564008567482e-05	2.53696998697706e-05	1.00258994102478	19.0443000793457	0.233816996216774	0.000159995004651137	3.00933006656123e-05	2.54182996286545e-05	0.991778016090393	18.389799118042	0.204215005040169	115000	-0.0024937700945884	0.00997505988925695	35	14	14	12	2	2	2	0.000114841997856274	7.19052013664623e-06	2.5483500957489	18.7497997283936	0.0679799988865852	0.000105869999970309	1.03619995570625e-05	4.58826017379761	18.8381004333496	0.106265999376774	9.80037002591416e-05	1.50958003359847e-05	19.9044990539551	18.9220008850098	0.167238995432854	0.000273990008281544	1.91638991964282e-05	2.15367999771843e-05	0.974824011325836	17.805700302124	0.0759399980306625	0.000295481993816793	2.98454997391673e-05	5.02192015119363e-05	1.00258994102478	17.7236995697021	0.109665997326374	0.000279867992503569	4.77376997878309e-05	0.00015027800691314	1.00093996524811	17.7826995849609	0.185196995735168	115000	-0.201290994882584	0.239612996578217

2. Clean the data#

As our interest is to classify the sources, we first “encode” the PS1-STRM classifications from string values to integers:

label_encoder = LabelEncoder()
tab['class_encode'] = label_encoder.fit_transform(tab['class'])

Next, we will select a subset of the table, specifically the “FPSFMag” magnitude values for the five PS1 bands (g, r, i z, and y). Following the strategy of Beck et al. 2020, we will normalize these magnitudes by subtracting the median. Then, we will impute missing values by setting them equal to the median. NOTE this may be a risky move! An alternative strategy is to simply remove the missing values from the training set, or to impose an N-\(\sigma\) limit.

bands = ['g', 'r', 'i', 'z', 'y']
tt = 'FPSFMag'

tags_f = np.ravel([[b+tt] for b in bands])
fluxes = [tab[tag] for tag in tags_f]
for k in range(len(tags_f)):
    fluxes[k] = fluxes[k] - np.median(fluxes[k]) 
    
    fluxes[k] = np.where(fluxes[k] < -100, np.median(fluxes[k]), fluxes[k])
    fluxes[k] = np.where(fluxes[k] > 100, np.median(fluxes[k]), fluxes[k])

In addition to measured fluxes, we will add galaxy colors to the sample. For the unique combinations of filters we will compute colors. As for the fluxes, we normalize by subtracting the medians, and then all missing values are set to the medians as above (but see note!).

combos = list(combinations(bands, 2))
colors = []
tags_c = []
for c in list(combos): 
    color = np.array(tab[c[0]+tt]-tab[c[1]+tt])
    
    color = color - np.median(color)
    color[color < -100] = np.median(color)
    color[color > 100] = np.median(color)

    colors.append(color)
    tags_c.append(c[0]+','+c[1]+','+tt)

We then combine the flux sample with the color sample to produce a 100000 x 15 catalog

sample = np.array(np.vstack([fluxes, colors]))
sample = np.array(sample).T

np.shape(sample)

(100000, 15)

3. Reduce dimensionality with PCA and tSNE#

To explore how our selected features recover the classifications from PS1-STRM, we will apply PCA to the full sample.

pca_sample = PCA(n_components=2).fit_transform(sample)
print(pca_sample.shape)

(100000, 2)

The shape of the reduced PCA sample is now 100000 x 2. To visualize the results, we plot the reduced sample and over-plot the populations for each of the four PS1 classes:

classes = ['GALAXY', 'QSO', 'STAR', 'UNSURE']

xmin, xmax = -6, 16
ymin, ymax = -6, 4

# Peform the kernel density estimate
xx, yy = np.mgrid[xmin:xmax:100j, ymin:ymax:100j]
positions = np.vstack([xx.ravel(), yy.ravel()])

fig = plt.figure(figsize=(10, 10))
colors = ['dodgerblue', 'orange', 'limegreen', 'magenta']
for k, c in enumerate(classes):
    spots = tab['class'] == c
    
    plt.subplot(2, 2, k+1)
    plt.scatter(pca_sample[:, 0], pca_sample[:, 1], s=2, marker='.', color='grey', alpha=0.01)# bins='log', cmap='Greys', alpha=0.5) 
    
    values = np.vstack([pca_sample[:, 0][spots], pca_sample[:, 1][spots]])
    kernel = st.gaussian_kde(values)
    f = np.reshape(kernel(positions).T, xx.shape)
    cntr = plt.contour(xx, yy, f, colors=colors[k])
    h1, _ = cntr.legend_elements()
    
    plt.legend([h1[0]], [c])
    plt.xlim(-7, 16)
    plt.ylim(-6, 6)
    plt.xlabel('PCA 1')
    plt.ylabel('PCA 2')
    
plt.show()

../../../_images/Classifying_PanSTARRS_sources_with_unsupervised_learning_24_0.png

By inspection of the above panels, it looks like the different classes occupy unique regions of the simple PCA parameter space, with considerable overlap. As a next test, we will use tSNE to reduce dimensionality of the original dataset. In this example we will randomly subset the sample by 10% so that the tSNE algorithm runs quickly:

%%time
idx = np.random.choice(sample.shape[0], 10000)
tsne_sample = TSNE(n_components=2, init='pca', perplexity=10, learning_rate = 10).fit_transform(sample[idx])

/usr/share/miniconda/lib/python3.9/site-packages/sklearn/manifold/_t_sne.py:996: FutureWarning: The PCA initialization in TSNE will change to have the standard deviation of PC1 equal to 1e-4 in 1.2. This will ensure better convergence.
  warnings.warn(

CPU times: user 1min 35s, sys: 106 ms, total: 1min 35s
Wall time: 52 s

And similar to the PCA analysis, we will visualize the results:

xmin, xmax = -45, 45
ymin, ymax = -45, 45

# Peform the kernel density estimate
xx, yy = np.mgrid[xmin:xmax:100j, ymin:ymax:100j]
positions = np.vstack([xx.ravel(), yy.ravel()])

fig = plt.figure(figsize=(10, 10))

for k, c in enumerate(classes):
    spots = tab['class'][idx] == c
    plt.subplot(2, 2, k+1)
    plt.scatter(tsne_sample[:, 0], tsne_sample[:, 1], c='black', s=2, alpha=0.01)
    
    values = np.vstack([tsne_sample[:, 0][spots], tsne_sample[:, 1][spots]])
    kernel = st.gaussian_kde(values)
    f = np.reshape(kernel(positions).T, xx.shape)
    cntr = plt.contour(xx, yy, f, colors=colors[k]) 
    h1, _ = cntr.legend_elements()
    
    plt.legend([h1[0]], [c])
    plt.title(c)
    plt.xlabel('tSNE Dimension 1')
    plt.ylabel('tSNE Dimension 2')
plt.show()

../../../_images/Classifying_PanSTARRS_sources_with_unsupervised_learning_28_0.png

Again, it looks like each class occupies a distinct region of tSNE parameter space, although it is a bit noisy with lots of overlap, especially between the “galaxy” and “unsure” classes.

4. Perform unsupervised classification#

Next we will perform unsupervised classification with clustering. There are many choices of algorithms (see the sklearn clustering documentation), and here we will apply the k-means algorithm, which requires that we specify the number of clusters a priori (4 in this case):

First, we will split the sample into training and test sets:

labels = tab['class_encode']

# Split off 30% of the data for testing
X_train, X_test, y_train, y_test, idx_train, idx_test = train_test_split(sample, labels, np.arange(len(labels)), test_size=0.3, random_state=42, shuffle=True)

kmeans = KMeans(init="k-means++", n_clusters=4)
kmeans.fit(X_train)

KMeans(n_clusters=4)

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

To visualize the results of the unsupervised clustering model, we will use the PCA representation of the test dataset and color the points by their labels assigned by k-means:

fig = plt.figure(figsize=(5, 5))

xmin, xmax = -10, 20
ymin, ymax = -10, 20

# Set up a grid for the kernel density estimate
xx, yy = np.mgrid[xmin:xmax:100j, ymin:ymax:100j]
positions = np.vstack([xx.ravel(), yy.ravel()])

pca_sample_test = pca_sample[idx_test]
kmeans_labels = kmeans.predict(X_test)

for k in range(4):
    pca_cluster = pca_sample_test[kmeans_labels == k]
    
    values = np.vstack([pca_cluster[:, 0], pca_cluster[:, 1]])
    kernel = st.gaussian_kde(values)
    f = np.reshape(kernel(positions).T, xx.shape)
    cntr = plt.contour(xx, yy, f, colors=colors[k], levels = np.linspace(0.01,0.12,6))

plt.title("Unsupervised Clustering\nk-means")
plt.xlabel("PCA 1")
plt.ylabel("PCA 2")
plt.xlim(-7, 16)
plt.ylim(-6, 6)
plt.show()

../../../_images/Classifying_PanSTARRS_sources_with_unsupervised_learning_36_0.png

When given the number of clusters, the k-means algorithm does a reasonable job of identifying distinct clusters. However, comparing these results with the true locations of the “galaxy”, “qso”, “star” and “unsure” classes in the PCA figure in Section 3, we can see that there is not a perfect correspondence between them.

5. Perform supervised clustering#

Given the class labels from the CNN analysis by Beck et al. 2020, we will also apply supervised classification (i.e., training the algorithm using a sample of known answers). Here we’ll use stochastic gradient descent (SGD).

Initialize the SGD classifier via sklearn. Here we will use the default parameters, but please see the SGDClassifier() documentation for more information on the available options.

%%time
sgd = SGDClassifier()

sgd.fit(X_train, y_train)

CPU times: user 1.15 s, sys: 0 ns, total: 1.15 s
Wall time: 1.15 s

SGDClassifier()

To visualize the results of the supervised classification model, we will use the PCA representation of the test dataset and color the points by their labels assigned by SGD:

fig = plt.figure(figsize=(5, 5))

xmin, xmax = -10, 20
ymin, ymax = -10, 20

# Peform the kernel density estimate
xx, yy = np.mgrid[xmin:xmax:100j, ymin:ymax:100j]
positions = np.vstack([xx.ravel(), yy.ravel()])

pca_sample_test = pca_sample[idx_test]
sgd_labels = sgd.predict(X_test)

for k in range(4):
    pca_cluster = pca_sample_test[sgd_labels == k]
    
    values = np.vstack([pca_cluster[:, 0], pca_cluster[:, 1]])
    kernel = st.gaussian_kde(values) 
    f = np.reshape(kernel(positions).T, xx.shape)
    cntr = plt.contour(xx, yy, f, colors=colors[k], levels = np.linspace(0.01,0.12,6))


plt.title("Supervised Classification\nSGD")
plt.xlabel("PCA 1")
plt.ylabel("PCA 2")
plt.xlim(-7, 16)
plt.ylim(-6, 6)
plt.show() 

../../../_images/Classifying_PanSTARRS_sources_with_unsupervised_learning_43_0.png

Unlike the unsupervised k-means classifier, the supervised SGD classifier appears to do a better job of identifying the true classifications (see Section 3), especially in the case of overlapping clusters in the PCA representation.

6. Assess the model performance#

To visulize the performance of the supervised and unsupervised model results, we will use a confusion matrix.

def plot_confusion_matrix(predictions, input_data, input_labels):
    
    # Compute the confusion matrix by comparing the test labels (ds.test_labels) with the test predictions
    cm = metrics.confusion_matrix(input_labels, predictions, labels=[0, 1, 2, 3])
    cm = cm.astype('float')

    # Normalize the confusion matrix results. 
    cm_norm = cm / cm.sum(axis=1)[:, np.newaxis]
    
    # Plotting
    fig = plt.figure()
    ax = fig.add_subplot(111)

    ax.matshow(cm_norm)

    plt.title('Confusion matrix', y=1.08)
    
    ax.set_xticks([0, 1, 2, 3])
    ax.set_xticklabels(['Galaxy', 'QSO', 'Star', 'Unsure'])
    
    ax.set_yticks([0, 1, 2, 3])
    ax.set_yticklabels(['Galaxy', 'QSO', 'Star', 'Unsure'])

    plt.xlabel('Predicted')
    plt.ylabel('True')

    fmt = '.2f'
    thresh = cm_norm.max() / 2.
    for i in range(cm_norm.shape[0]):
        for j in range(cm_norm.shape[1]):
            ax.text(j, i, format(cm_norm[i, j], fmt), 
                    ha="center", va="center", 
                    color="white" if cm_norm[i, j] < thresh else "black")
    plt.show()

plot_confusion_matrix(kmeans_labels, X_test, y_test)

../../../_images/Classifying_PanSTARRS_sources_with_unsupervised_learning_48_0.png

From the above plot, it is clear that the unsupervised clustering algorithm does an extremely poor job of identifying the correct classes!

plot_confusion_matrix(sgd_labels, X_test, y_test)

../../../_images/Classifying_PanSTARRS_sources_with_unsupervised_learning_50_0.png

Based on the above confusion matrix, the SGD classifier performs very well at identifying stars (~94% true positive) whereas it does not perform as well at distinguishing between the galaxy, QSO and unsure classes.

FAQs#

How do I interpret these results? The analysis above indicates that meaningful information about unresolved astronomical sources can be extracted from their simple measured fluxes and colors. In particular, the SGD classifier is able to identify stars well, even with the out-of-the-box parameters from sklearn. However, unsupervised k-means clustering is not able to successfully classify sources.
Can I improve the model by changing it? Yes! In the k-means clustering and SGD classifier cases, there are many model parameters which can be tuned to improve the performance of the model. For example, in SGD, you can select the loss function (here we use the default, “hinge” which gives a linear support vector machine.
Can I try a different model? I think the results could be improved. Yes! Check out the documentation at sklearn about classification. In particular, the sklearn flowchart is a useful tool for figuring out which (of the many!) options might be useful.
I want to test my model on my training data! No. You will convince yourself that your results are much better than they actually are. Always keep your training, validation, and testing sets completely separate!

Appendix#

How to extract the subset of data analyzed here from the PS1 and PS1-STRM catalogs

Instead of working with the full PS1-STRM catalog, we extracted a subset of 100,000 sources.

First, we downloaded a single PS1-STRM file (all sources with Declinations between 77 and 90) from MAST. We then randomly selected objects so that each of the four classes (“galaxy”, “qso”, “star”, “unsure”) had equal representation in a sample of 100,000 objects. We then saved the object IDs and unique internal object identifiers to a csv catalog file.

Then, we uploaded this file to MAST CasJobs as “MyTable”. We first cross-matched the object IDs we selected with the PS1-STRM catalog via the following query:

select t.objid, s.class, s.prob_Galaxy, s.prob_Star, s.prob_QSO, s.z_phot, s.z_photErr
into MyDB.ps1strm_fmo1
from MyDB.MyTable t
join catalogRecordRowStore s on s.objid=t.objid and s.uniquePspsOBid=t.uniquePspsOBid

We then cross-matched this catalog with the full PS1 ForcedMeanObject table with the following query:

select s.class, s.prob_Galaxy, s.prob_Star, s.prob_QSO, s.z_phot, s.z_photErr, fmo.
into MyDB.ps1strm_fmo
from MyDB.ps1strm_fmo1 s
join ForcedMeanObject fmo on fmo.objid=s.objid

About this Notebook#

Author:
Claire Murray, Assistant Astronomer, cmurray1@stsci.edu

Info:
This notebook is based on the PS1-STRM catalog (Beck et al. 2020).

Updated On: 2022-5-25

Citations#

If you use this data set, astropy, or sklearn for published research, please cite the authors. Follow these links for more information:

Top of Page Space Telescope Logo

Hello Universe!

Classifying Pan-STARRS sources with unsupervised and supervised learning

Contents