SDPipelineHeuristicsStatusReport.pptVIP

ALMA Pipeline Heuristics F2F Meeting 2006 in Paris SD Pipeline Heuristics Status Report ALMA Project, NAOJ George KOSUGI E2E Pipeline Process Flow Current SD Pipeline Flow Chart Environment for SD Heuristics Pipeline Python Packages casapy with ASAP numarray: work area for spectrum data analysis Scipy: Kmean clustering algorism Numpy: data for Kmean PyLab: plotting PyFits: for collecting info from SDFITS header What is necessary for SD UT MS input Need sample MS MSs for casa/ASAP UT in 2007/03 will be adequate Store/Retrieve info into/from AIPS++Table MS and FITS output ----------------- 2~3 FTE-w Calibration script Investigate TelCal Engine Investigate scripts used in the casa/ASAP UT ----------------- 2~3 FTE-w ? Additional plots ----------------- 1~2 FTE-w ? Baseline Fit is done for each spectrum one by one Determine Line Free Channels (if not specified) Fit and Subtract Baseline Calculated from the Line Free Channels Determine Line Free Channels (if not specified) Subtract DC (bias) and gradient components Accumulate several spectra in close vicinity in space to enhance S/N ratio for line detection Line Detection by application of ASAP.asaplinefinder Kmean clustering analysis is applied to the detected lines (parameters are line center and line width) Select real (supposed to be) lines by examining continuities in space, distortion of the vector, density of the detection, and cluster size Calculate 2-dimensional fit (ra, dec, line center), (ra, dec, line width) to each cluster member Apply fit to spectra whose position is within the clustering radius Spectrum based Flagging (not channel based) Statistics are measured for each spectrum (emission line channels are excluded): baseline rms before baseline subtraction after baseline subtraction difference from the running mean before baseline subtraction after baseline subtraction may need other criteria Tsys, etc... Plot statistics Gridding (Align to Lattice pattern) Beam size and Grid spacing are given Spectr