Differences

This shows you the differences between two versions of the page.

--- j2ms2_and_tconvert [2019/02/25 15:49] – verkouter
+++ j2ms2_and_tconvert [2019/02/28 12:26] (current) – benito
@@ Line 1: / Line 1: @@
-== j2ms2 and tConvert ==
+====== j2ms2 and tConvert ======
 The normal workflow of processing correlator data at JIVE is to take the correlator output (SFXC, UniBoard) and:
-  - use ''j2ms2'' to convert to CASA MeasurementSet format
+  * use ''j2ms2'' to convert to CASA MeasurementSet format
-  - inspect the data, do flagging, other stuff ...
+  * inspect the data, do flagging, other stuff ...
-  - convert MeasurementSet data to FITS-IDI format using ''tConvert'' for archiving and distribution to the PI
+  * convert MeasurementSet data to FITS-IDI format using ''tConvert'' for archiving and distribution to the PI
 These tools are currently undergoing a big overhaul; modernizing the build system, changing the project layout from a CASA/AIPS++ subproject into a standalone one, migrating revision control system from CVS to git and, most importantly, addressing built up changes to the JIVE SFXC correlator capabilities.
@@ Line 21: / Line 21: @@
 ''/eee/bin/j2ms2'' and ''/eee/bin/tConvert''
-Features:
+Features ("bugs" in this context):
-- somewhat broken ''tConvert''; may lose up to a second of data between IDI-chunks for normal experiments, can be a lot worse for mixed-mode experiments where a significant number of stations drop a number of subbands
+  * somewhat broken ''tConvert''; may lose up to a second of data between IDI-chunks for normal experiments, can be a lot worse for mixed-mode experiments where a significant number of stations drop a number of subbands
-- leaks metadata from multiple-experiments-per-single-vex-file into individual experiments' MeasurementSet (and consequently in the FITS-IDI files).
+  * leaks metadata from multiple-experiments-per-single-vex-file into individual experiments' MeasurementSet (and consequently in the FITS-IDI files).
-- ''j2ms2'': if for some reason a baseline is to be written in reverse order then the spectrum would not be flipped
+  * ''j2ms2'': if for some reason a baseline is to be written in reverse order then the spectrum would not be flipped
@@ Line 33: / Line 33: @@
 after which the binaries can be found here:
-''/eee/src/jive-casa/code/jive/bin/x86_64/{j2ms2, tConvert}''
+''/eee/src/jive-casa/code/jive/bin/x86_64-64/{j2ms2, tConvert}''
 Features:
-- do not leak meta data; only meta data for antennas, frequencies and sources that have data in the MeasurementSet will be written. The filtering of the antenna table is off by default (ie it will leak all antenna data into all measurement sets) but can be enabled by adding ''fo:squash_antenna_table'' on the ''j2ms2'' command line.
+  * do not leak meta data; only meta data for antennas, frequencies and sources that have data in the MeasurementSet will be written. The filtering of the antenna table is off by default (ie it will leak all antenna data into all measurement sets) but can be enabled by adding ''fo:squash_antenna_table'' on the ''j2ms2'' command line.
-- baseline data is now written correctly by ''j2ms2'' even when baselines are written in reverse order in the MS
+  * baseline data is now written correctly by ''j2ms2'' even when baselines are written in reverse order in the MS