QSO:::BAL have a high failure rate

[londumas]

Most of the failures of quasars in redrock+archetype are linked to BALs.
Regular QSO have a near 97% good rate and 1.5% failure rate, whereas BAL QSO have a 82% good rate and 13.3% failure rate.
The current PCAs and archetypes don't have BAL in them.
Currently desisim.SIMQSO can't be used to produce neither PCA nor archetypes and don't have BALs.

Using DR12_superSet:

All

                   ntarg   good  fail  miss  lost
GALAXY             22788  87.85 10.27  0.44  1.44
QSO               295310  95.43  2.63  0.59  1.34
STAR              205250  88.96  2.10  5.51  3.44

                         | GALAXY  | QSO     | STAR    | 
GALAXY             22360 |  95.21 |   3.70 |   1.08 | 
QSO               289597 |   1.31 |  98.57 |   0.11 | 
STAR              186894 |   1.79 |   0.57 |  97.65 |

No BAL

                   ntarg   good  fail  miss  lost
GALAXY             22788  87.85 10.27  0.44  1.44
QSO               266058  96.89  1.47  0.58  1.06
STAR              205250  88.96  2.10  5.51  3.44

                         | GALAXY  | QSO     | STAR    | 
GALAXY             22360 |  95.21 |   3.70 |   1.08 | 
QSO               261701 |   1.10 |  98.81 |   0.09 | 
STAR              186894 |   1.79 |   0.57 |  97.65 | 

Only BAL

                   ntarg   good  fail  miss  lost
QSO                29252  82.13 13.23  0.68  3.95

                         | GALAXY  | QSO     | STAR    | 
QSO                27896 |   3.29 |  96.38 |   0.33 |

[andreufont]

Hi Helion, These type of tests are very important. Jaime, Luz Angela and Alma (added in this email) have looked at adding BALs in SIMQSO, but I'm not sure about the status of this effort. Any update? Best, Andreu

…

On Tue, 26 Jun 2018, 23:08 Hélion du Mas des Bourboux, < ***@***.***> wrote: Most of the failures of quasars in redrock+archetype are linked to BALs. Regular QSO have a near 97% good rate and 1.5% failure rate, whereas BAL QSO have a 82% good rate and 13.3% failure rate. The current PCAs and archetypes don't have BAL in them. Currently desisim.SIMQSO can't be used to produce neither PCA nor archetypes and don't have BALs. Using DR12_superSet: All ntarg good fail miss lost GALAXY 22788 87.85 10.27 0.44 1.44 QSO 295310 95.43 2.63 0.59 1.34 STAR 205250 88.96 2.10 5.51 3.44 | GALAXY | QSO | STAR | GALAXY 22360 | 95.21 | 3.70 | 1.08 | QSO 289597 | 1.31 | 98.57 | 0.11 | STAR 186894 | 1.79 | 0.57 | 97.65 | No BAL ntarg good fail miss lost GALAXY 22788 87.85 10.27 0.44 1.44 QSO 266058 96.89 1.47 0.58 1.06 STAR 205250 88.96 2.10 5.51 3.44 | GALAXY | QSO | STAR | GALAXY 22360 | 95.21 | 3.70 | 1.08 | QSO 261701 | 1.10 | 98.81 | 0.09 | STAR 186894 | 1.79 | 0.57 | 97.65 | Only BAL ntarg good fail miss lost QSO 29252 82.13 13.23 0.68 3.95 | GALAXY | QSO | STAR | QSO 27896 | 3.29 | 96.38 | 0.33 | — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#129>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ADscR0QHIvhkowTUu5khifVz_kCf1UP7ks5uArDHgaJpZM4U4wUr> .

[forero]

@andreufont Luz is quickly ramping up on python+desihub. We hope to have a PR to discuss by the lya telecon next week.

[moustakas]

This is a known issue but we did reasonably well for the Y1 analyses, but we can always do better. Closing this ticket but let's revisit this question once the new set of Y3 (see #280) templates are in place.