Skip to content

Accellerate calculation of standard deviations #261

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
tammojan merged 2 commits into from
Aug 5, 2025
Merged

Accellerate calculation of standard deviations #261

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1d97097
Save on standard deviation calculation
Jul 30, 2025
34070da
Adapt also the last sigma calculation
Jul 30, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 9 additions & 3 deletions bdsf/functions.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2143,7 +2143,8 @@ def bstat(indata, mask, kappa_npixbeam):
maxiter = 200
converge_num = 1e-6
m_raw = numpy.mean(skpix)
r_raw = numpy.std(skpix, ddof=1)
d2 = (skpix-m_raw)**2
r_raw = (d2.sum()/(ct-1))**0.5

while (c1 >= c2) and (iter < maxiter):
npix = skpix.size
Expand All @@ -2156,21 +2157,26 @@ def bstat(indata, mask, kappa_npixbeam):
kappa = 3.0
lastct = ct
medval = skpix[npix//2]
sig = numpy.std(skpix)
if iter == 0:
sig = r_raw
else:
m_new = numpy.mean(skpix)
sig = (d2.sum()/ct - (m_raw-m_new)**2)**0.5

wsm1, wsm2 = numpy.searchsorted(skpix, [medval - kappa*sig,
medval + kappa*sig])
ct = wsm2 - wsm1
if ct > 0:
skpix = skpix[wsm1:wsm2]
d2 = d2[wsm1:wsm2]

c1 = abs(ct - lastct)
c2 = converge_num * lastct
iter += 1

mean = numpy.mean(skpix)
median = skpix[skpix.size//2]
sigma = numpy.std(skpix, ddof=1)
sigma = (d2.sum()/(ct-1) - (m_raw-mean)**2 * ct/(ct-1) )**0.5
mode = 2.5*median - 1.5*mean
if sigma > 0.0:
skew_par = abs(mean - median)/sigma
Expand Down