Transcript Mohanty_DCR

DCR
Plan of presentation
• Soumya Mohanty: Overview, aims & work done
• R. Balasubramanian: Details of Hardware, Database
• Using the database:
• R. Balasubramanian
• Soma Mukherjee
• S. Babak
21 Feb 2002
Soumya D. Mohanty, AEI
1
What is DCR?
•Quality flags
•...
DATA MINING
All Channels
DCR
Database of
change pts.
EXPLORATORY
ANALYSIS
Short data
Stretches
21 Feb 2002
Soumya D. Mohanty, AEI
2
Using a database: Data Mining &
Data Exploration
• Different but complementary approaches.
• Data exploration:
• I want to see the time series corresponding to a bunch of triggers that I
selected from a database. (Then do more analysis on this selected data.)
• Typically, Follow up data is short, Quick look environment needed, no
specific queries
• Data Mining:
• Can the transients seen over a month be classified into groups? What was the
rate of transients in each group as a function of time (Maybe some types occur
in the day, some occur in the night). (Then use this information to quantify the
quality of long data stretches).
• Purely database based; Re-analysis of raw data may be impractical
21 Feb 2002
Soumya D. Mohanty, AEI
3
What is special about DCR?
Any such transformation
will introduce errors
Raw noisy data
Information
Transformer
DATABASE
21 Feb 2002
Soumya D. Mohanty, AEI
•Spurious information
•Misses genuine stuff
DCR: Control
the false alarm
rate
4
Control on False Alarm Rate
• Important for Data mining
• Statistical analysis done on database itself since reanalysis of long
stretch of data expensive
• Need to put error bars
• Not so important for Data exploration
• Looking for information about specific events
• Each explorer will work with his/her own short data stretch
21 Feb 2002
Soumya D. Mohanty, AEI
5
Initial Design of DCR
Soumya Mohanty, Soma Mukherjee, CQG, 2002.
Restricted
DCR (rDCR)
21 Feb 2002
Soumya D. Mohanty, AEI
6
rDCR project: Beginnings
• Group formed Sep’01
• AEI: Soumya Mohanty (lead), Soma Mukherjee
• Cardiff: R.Balasubramanian, S.Babak, D.Churches,
B.S.Sathyaprakash
• Birmingham: A.Vecchio, R.Ingley (new)
• Decided to include Kalman Filter also
• Made list of software components: GEO++ and
support
• Much thinking on the Design of GEO++
• Coding started Oct’01
21 Feb 2002
Soumya D. Mohanty, AEI
7
Aspects of rDCR project, Status
•
•
•
•
Building GEO++
Integration of GEO++, Frames, MPI, MySQL
Integration of entire chain on Hannover Beowulf
Implementing DCR algorithms
• MBLT: existed as a mixture of C and matlab codes
• PSDCD: exists as a C++ (DMT ready) code
• Simple Non-stationarity detector
• Database Analysis
• Administrative: CVS, Web Site, Coding conventions etc.
21 Feb 2002
Soumya D. Mohanty, AEI
8
GEO++
• Major component of all the work done so far
• All GEO++ codes planned have been completed
and tested
21 Feb 2002
Soumya D. Mohanty, AEI
9
Frame, MPI integration
• Reading Frames: adopted a simple model for data
•
•
•
•
requests
Status: Done
Purpose of MPI code: to encapsulate a GEO++
based code and run it on a Beowulf cluster
One slave – one channel dataflow model
implemented
Status: Done
21 Feb 2002
Soumya D. Mohanty, AEI
10
Integration with MySQL
• Creating an elegant and easy to use OO design
took most of the time
• Writing to database done
21 Feb 2002
Soumya D. Mohanty, AEI
11
Run entire chain on the
Hannover Beowulf
• Making Beowulf ready, installing software took
about a month
• Done
Get Data
21 Feb 2002
Read
Frames
Run Simple
monitor
on all
channels
Soumya D. Mohanty, AEI
Database
12
rDCR Algorithms:
Implementation
• MBLT code in C++ written
• Completion of MBLT was dependent on a number of lower
level classes
•
Filter, Filter Design, Demod, Modulate, Running Median.
• Code now complete and under testing
• Did not get a significant increase in performance over
Matlab code
• Mainly because the most time consuming part is coded in C and
the same is used here.
• This is quite preliminary
21 Feb 2002
Soumya D. Mohanty, AEI
13
Database Analysis
• A simple monitor was built to test IO and as
many classes as possible while the DCR
algorithms were being coded
• Main aim: start exercising the post-database
analysis chain (software, logistics)
• Software: Matlab, Triana
• Logistics: Need to transfer tables to local
machines
21 Feb 2002
Soumya D. Mohanty, AEI
14
DCR on the Web
http://www.aei.mpg.de/~mohanty/DCR/DCRindex.html
21 Feb 2002
Soumya D. Mohanty, AEI
15
Current Status & Plans
•
•
•
•
All software originally planned in place
MBLT code ready; testing
Database analysis exercises done
Plans:
•
•
•
•
•
•
Construct a simple non-stationarity test
Complete rDCR
Complete Kalman Filter
Run on all of coincidence run data
Extend GEO++
Analyse rDCR database
21 Feb 2002
Soumya D. Mohanty, AEI
16