Gregory Mendell and Mike Landry

1
The StackSlide Search
http//www.ligo-wa.caltech.edu
Summary November 2003 LSC Meeting

• Gregory Mendell and Mike Landry
• LIGO Hanford Observatory

2
The StackSlide Search

• An incoherent search method that stacks and
  slides power to search for periodic sources. (P.
  Brady T. Creighton Phys. Rev. D61 (2000)
  082001 gr-qc/9812014.)
• The periodic search is computationally bound. A
  hierarchical approach that combines coherent
  incoherent methods is needed to optimize
  sensitivity.
• Sources like LXMBs with short coherence times (
  2 weeks) require incoherent methods.
• Want to add StackSlide to the incoherent toolbag
  along with Hough transforms.
• Mike Landry and I proposed that we write a
  stackslide DSO in June 2003.

3
StackSlide ABCs
Bins with frequency domain data, e.g., from SFTs.
PSDs or F-statistic from SFTs
B. Make Stacks
C. Slide and output Sum
A. Input Blocks of data
(from coherent methods)
(incoherent step)
4
Progress to Date
We have working code that stack slides PSDs
from SFTs on a flat parameter space!
Implemented SLIDING since last LSC
5
StackSlide Basic Statistics
One stack
(N samples)
(Ave M stacks)
Detection if
For A2N gt ?2
For A2N lt ?2
M number of SFTs. If F-statistic P(x)
Cx2M-1e-Mx for mean F 2.
PDF for gaussian noise, normalized to mean of 1.
6
Study of E10 Injected Pulsar 0.
7
Study of E10 Injected Pulsar 2.
8
Study of E10 Injected Pulsar 3.
9
Study of E10 Injected Pulsar 4.
10
Big to do list still exists

• Next major effort is to debug on simple test data
  and S2/E10/S3 injections.
• Need to add code that handles thresholds,
  parameter space metric, stacks F-statistic, runs
  in parallel, etc.
• Need to decide on windowing/overlapping/oversampli
  ng of SFTs etc
• SET UP PIPELINE THAT LEAD TO DETECTION OR UPPER
  LIMIT!

Goal

• Set preliminary upper limit on S2/S3 data on same
  parameter space as Hough code.