Upper Limit calculation with the last enrico's version

Manuel Castro Avila

unread,

Feb 15, 2017, 11:55:21 AM2/15/17

to enrico

Hi,

I have downloaded the last ernico. When a run enrico_lc and upper limit calculations is requiered I get an error message:

# ************************************************************
# *** 14 UpperLimit --- Compute upper Limit
# ************************************************************
[FitMaker]: Assumed index is 1.5
Parameter Index not found for source J0109.8+6132 in file /home/castro/Fermi/Sources/Rodrigo/J0109/LightCurve_417bins//J0109.8+6132_PowerLaw_weekly_LC_0_out.xml

Also for some reason, in the .xml file, the source model uses EblAtten::LogParabola instead PowerLaw

<source name="J0153.4+7114" type="PointSource">
    <spectrum type="EblAtten::LogParabola">
      <parameter free="1" max="2.5" min="0" name="tau_norm" scale="1" value="1"/>
      <parameter free="0" max="5" min="0" name="redshift" scale="1" value="0.022"/>
      <parameter free="0" max="20" min="0" name="ebl_model" scale="1" value="4"/>
      <parameter free="1" max="1000" min="1e-05" name="norm" scale="1e-15" value="2.31896"/>
      <parameter free="1" max="5" min="0.5" name="alpha" scale="1" value="1.56702"/>
      <parameter free="0" max="300000" min="30" name="Eb" scale="1" value="300"/>
      <parameter free="0" max="0" min="0" name="beta" scale="1" value="0"/>
    </spectrum>
    <spatialModel type="SkyDirFunction">
      <parameter free="0" max="360" min="-360" name="RA" scale="1" value="28.3577"/>
      <parameter free="0" max="90" min="-90" name="DEC" scale="1" value="71.2518"/>
    </spatialModel>
</source>

In a previous enrico's version from mid-last year, Upper limit calculation worked and the source was fitted by using a PowerLaw. Also I noticed an additional step in the last enrico's version:

# ************************************************************
# ***   1   gtselect --- Select data from library, coarse cut
# ************************************************************

Please, someone can explain me what does it mean?

Thanks in advance,

Best regards,

Manuel

David Sanchez

unread,

Feb 16, 2017, 2:42:56 AM2/16/17

to enrico

Hi Manuel

thanks for using enrico. UL calculation works only for PL since there is no point to make an UL on a more complexe model.

cheers

david

Manuel Castro Avila

unread,

Feb 16, 2017, 8:55:42 AM2/16/17

to enrico

Hi David,

Thanks for the prompt reply.

So how can I force that the source modeling be performed with PL instead EblAtten::LogParabola when I run enrico_lc?. I have set in the .conf file spectrum=Powerlaw, I don't if it is only for enrico_sed?

In a previous enrico's version the model used for enrico_lc was PowerLaw, but it the last version the model is set to EblAtten::LogParabola by default. Is there any way to set the model to PL when the .xml file is generated?

Best regards,

Manuel

David Sanchez

unread,

Feb 16, 2017, 9:02:21 AM2/16/17

to gammapy...@googlegroups.com

Did you generate the .xml file after the change of the spectrum to Powerlaw?

best

david

--
You received this message because you are subscribed to a topic in the Google Groups "enrico" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/gammapy_enrico/G_pN2E0jIbw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to gammapy_enrico+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

David Sanchez

Manuel Castro Avila

unread,

Feb 16, 2017, 9:21:30 AM2/16/17

to enrico

Hi David,

Yes, first I generate the .conf file with enrico_conf.

[target]
        name = J0109.8+6132
        ra = 17.4431
        dec = 61.5585
        redshift = 0.783
        ebl_model = 4
        spectrum = PowerLaw
        # 0=Kneiske, 1=Primack05, 2=Kneiske_HighUV, 3=Stecker05, 4=Franceschini, 5=Finke, 6=Gilmore, 10=
        # full list http://www-glast.stanford.edu/cgi-bin/viewcvs/celestialSources/eblAtten/eblAtten/Ebl
        fit_tau = yes

and after I run enrico_xml:

<source name="J0109.8+6132" type="PointSource">

The source's model is generated to use EblAtten::LogParabola,

To unsubscribe from this group and all its topics, send an email to gammapy_enric...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
David Sanchez

David Sanchez

unread,

Feb 16, 2017, 9:26:41 AM2/16/17

to gammapy...@googlegroups.com

if you put 0 in redshift this will solve the problem. if the redshift is not zero, EBL attenuated model is used.

cheers

davud

To unsubscribe from this group and all its topics, send an email to gammapy_enrico+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

David Sanchez

Reply all

Reply to author

Forward