2001 Visiting Committee Report AIPS J. Ulvestad April 2001 We have resumed the practice of releasing frozen versions of AIPS, with a 31DEC00 version. Hereafter, we intend to make a release once annually. This practice has been necessitated by the fact that most of our external users have been unwilling to run the regular midnight job to get the test version of AIPS (31DEC01) that is updated daily. (At this writing, only about 15 non-NRAO sites are actively running the midnight job to acquire 31DEC01.) A significant fraction of the time spent responding to bug reports was being wasted in trying to figure out which versions of various programs people were using, so some frozen release was needed to reduce this. Of course, doing a full release also takes a great deal of time, so one release per year seemed to be the best compromise. As reported last year, between 70% and 75% of AIPS distributions are received by ftp; many CD-ROM versions are apparently distributed for use on laptop computers running Linux. Two AIPS memos have been produced over the last year. The first is Memo No. 104, by Eric Greisen, entitled "Y2K, a new DDT, and AIPSMark(00) Measurements." The second is Memo No. 105, by Jim Ulvestad and Eric Greisen, entitled "AIPS Procedures for Initial VLBA Data Reduction." In addition, VLBA Scientific Memo No. 25 is relevant. This memo, by Jim Ulvestad, is entitled "A Step-by-Step Recipe for VLBA Data Reduction in AIPS," and has also been added to the AIPS Cookbook as Appendix C. These last two memos are being updated regularly as new procedures are added. In May 2000, Jim Ulvestad took over the nominal supervision of the AIPS group. Eric Greisen made an extended visit to Socorro from April 2000 through October 2000, and then moved to Socorro permanently. Following Ketan Desai's departure in April 2000, Chris Flatters left the AIPS group in August 2000 to work in the computing industry, after producing an initial set of VLBA data-reduction scripts. In January 2001, Amy Mioduszewski joined the group; her primary role has been to expand the VLBA scripts and to provide support to users reporting VLBI bugs/problems to the designated AIP mailing list (the rotation of designated AIPS has been discontinued due to lack of personnel).. Since the beginning of 2000, the AIPS group has lost two experienced programmers, and is largely limited to bug-fixing and very limited new developments. Currently, the group consists of three scientific staff members in Socorro, plus three other individuals who spend a small fraction of their time on AIPS. The net commitment to AIPS is between 2.5 and 3 full-time-equivalent employees. At this writing, large file support (greater than 2 Gbytes) has been tested and is is being implemented under version 2.4.2 of the Linux kernel; we expect this to be of great benefit to users with large spectral- line data sets. Other highlights from 2000/2001 are listed below; potential users are reminded that many of these features are available only in the 31DEC01 and (possibly) 31DEC00 versions of AIPS. New Features in AIPS during 2000/2001 Task SCIMG was added to iteratively image and self-calibrate a multi-field data set. Task SPFLG was revised to improve editing of spectral-line data. FILLM and all calibration tasks have been modified to enable antenna weights to be calibrated, as described in AIPS Memo No. 103. This can significantly reduce RMS noise for data sets in which antennas have substantially different sensitivities (e.g., 22 and 43 GHz on the VLA). Task FGPLT was added to display the times when data samples are flagged. Tasks UVCON and CONFI have been improved substantially to support ALMA configuration studies. Most chapters of the AIPS Cookbook have been modified, with particular attention paid to Chapter 9, on VLBI data reduction. IMAGR has been modified to improve speed and flexibility when Cleaning. FACES was added to input catalogues (e.g., NVSS) to model the sky for multi-field imaging. FLATN now does mosaicing as well as multiple fields. SETFC was improved to determine a better pattern of "faces" needed to image the primary beam area fully. Task PBEAM has been added to fit primary-beam power patterns, and new beam patterns have been inserted in other tasks at all VLA wavelengths. FXPOL was rewritten to enable more flexible correction of polarization labeling for data from all VLBI correlators. A new RUN file VLBAUTIL has been added to the system in order to define procedures to simplify VLBA data reduction. The procedures currently cover all steps from data loading through fringe-fitting and phase-referencing. More details are in AIPS Memo No. 105. The CROSSPOL script for calibrating VLBI polarization leakage was made more robust. FITLD was modified to better cope with non-VLBA correlators UVFLG can now flag data above or below a specified elevation, or within an elevation range. UVPLT now uses REFANT to plot values such as elevation angle for individual antennas. MATCH matches sources, antennas, and frequencies from multiple VLBI data sets. XAS was modified to work on 24-bit true color displays. The new Y2K test suite was developed to provide a more demanding set of tests for the current generation of faster computers. See AIPS Memo No. 104 for details. The VLA flux scale was updated in SETJY, with values for 1999.2 added. The AIPS installation wizard was finished. Significant Bugs Resolved CALIB and FRING were rounding times outward by too much, causing confusion about which data and which scans were which. FITAB had an error that occasionally destroyed the first part of input uv data files. BPASS was not shifting spectra properly for the cross-correlation case. Plotting programs changed to fix coordinate errors when plotting around the poles. Goals for 2001/2002 Continue maintenance. Improve user support. Support midnight jobs and installations at all NRAO sites. Continue to streamline VLBI data-reduction procedures, and update related documentation. Freeze and release 31DEC01 near the end of 2001. Low-level code development supporting NRAO instruments. Response to 2000 Visiting Committee Report and Recommendations AIPS enhancements: AIPS developments and enhancements crucial to NRAO instruments have continued. Several important cases are the new data-weighting schemes, the development of streamlined procedures for VLBA data reduction, and tasks for configuration simulations for ALMA.