scantable.auto_quotient fails on specific data set

[erik.muller@…]

ASAP segv when

s=scantable("2006-09-03_1944-M165.rpf")
q=s.auto_quotient()

The data is on /DATA/KAPUTAR_3/mopra/MOPSarchive/

[Malte Marquarding]

Acknowledged as a bug.

Just a few little hints:

quot=s.auto_quotient

means 'quot' is an alias for 's.auto_quotient'

quot=s.auto_quotient()

means 'quot' holds the result of the function 's.auto_quotient'

Always use brackets, unless you know better ;-)

[Malte Marquarding]

Work around - manual quotient:

s = scantable("2006-09-03_1944-M165.rpf")
sel = selector()
# ons
sel.set_query("SRCTYPE==0")
s.set_selection(sel)
son = s.copy()

s.set_selection()

#offs
sel.set_query("SRCTYPE==1")
s.set_selection(sel)
soff = s.copy()

q = quotient(son,soff)

I'll investigate why it crashes.

[Malte Marquarding]

Resolved

I was (foolishly) assuming the timestamp within a cycle (e.g. for all IFs within a cycle to be the same). This is not the case for this data set (probably rounding). I am now looking within 0.5s of the minimum of delta t. As cycles are 2s or longer this should be ok.

[Malte Marquarding]

This fix seem to have opened up another issue on the following data set.

http://www.atnf.csiro.au/people/Jim.Lovell/t015_2005_054.tar.gz

s=scantable('2005-02-23_090305_T015.rpf')

q=s.auto_quotient()

ngc3620=q.get_scan('NGC3620')

ngc3620.auto_poly_baseline(order=1)

ngc3620av=ngc3620.average_time(weight='tsys',align='true')

[Malte Marquarding]

I think this was fixed long ago. New ticket if it happens again.