Data Analysis Plan

1
Data Analysis Plan

• GBT IDL Workshop
• October 15, 2004

2
Meeting Outline

• Morning activities
• Generate a list of discussion items
• Demonstrations
• Lunch
• Discussion forum
• Dinner
• Summary

3
Science Drivers

• Interactive package for basic single-dish
  analysis
• Available at NRAO and home institutions
• Import/Export data formats
• Well-documented data model (data structure and
  data flow)
• Real-time data checks / post-processing
• Handle our diverse observing modes, frequencies,
  backends, .
• Handle new instruments and analysis techniques
• Pipelining, cameras, etc.

4
Expected Customers

• Black belts
• Unique experiments
• Explore every step in the analysis
• Can develop new algorithms
• Can share these algorithms
• Novices
• Use canned procedures developed by others
• Pipelining

5
Immediate Goals of the Data-Handling Project

• Interactive package for BASIC analysis
• Leverage observers past experiences
• Allow observers to contribute to the development
  of enhancements
• Sharing of contributions
• Short time scale for a significant improvement
• Limited goals for what NRAO contributes
• Rely on other systems for other analysis needs

6
Options

• Resurrect UniPOPS (argh!)
• Reinvest in AIPS/DISH or AIPS
• Go with a community standard
• Invest in CLASS, IRAF,
• Invest in a commercial package
• IDL
• MATLAB
• Mathematica

7
Why IDL? Resources/Familiarity

• IDL brings
• Large user base
• Extensive built-in libraries, algorithms,
  graphics
• Rapid development environment
• External users
• Tom Bania 3He package
• Robishaw Polarization
• Kondratko H2O Spectroscopy
• Other observatories
• FUSE, Spitzer, .
• IDL Astronomy Users Library
• Arecibos IDL routines
• Internal resources
• Langston IDL Real-Time data display
• ONeil IDL experience

8
Why a UniPOPS philosophy?

• Familiarity
• Simple syntax
• Interactive, efficient, .
• Emphasizes a toolbox of single-purpose analysis
  steps
• Exploratory analysis, new algorithms
• Scripting for generating complex analysis from
  these simple commands
• User extendable
• Data export to move data to other systems for
  more complicated analysis
• Online and Offline analysis

9
Extent of the IDL Project

• Wont Include (initially)
• Advanced imaging
• Pulsar, Radar, .
• Adding VLA data to GBT data
• Bolometer cameras
• Pipelining
• .
• Will include Infrastructure
• First priorities
• Data Model
• I/O, Data Base
• Toolbox of commands available from two different
  user levels

10
Available Resources Timeline

• Details from Nicole

11
IDL An integral part of the observatorys plans

• CLASS / AIPS
• FITS for data import/export
• ALMA
• Resource sharing -- Significant common
  infrastructure
• E2E
• Data pipelining using AIPS technology
• Learn from our IDL experiences and IDL Data
  models.
• OBIT
• Prototype analysis for the Penn Array
• AIPS/DISH
• Based on appropriate data model from IDL
  experience, OBIT, and ALMA
• Python instead of GLISH, CASA
• Combining VLA data with GBT data
• Advanced GBT imaging