|
|
| Changed: |
<
< |
- molecule, recombc, recombhe, recombh*, recombn*, and recombo had a bug that resulted in the marked lines some times appearing at the wrong position. This would occur when the plotted x-axis included an additional offset (usually the source velocity). The shift due to that offset was applied in the wrong direction.
|
>
> |
- molecule, recombc, recombhe, recombh, recombn, and recombo had a bug that resulted in the marked lines some times appearing at the wrong position. This would occur when the plotted x-axis included an additional offset (usually the source velocity). The shift due to that offset was applied in the wrong direction.
|
|
|
| Changed: |
<
< |
- Data selection by sampler name is now available in the following routines: flag, getbs, getfs, getnod, getps, getsigref and gettp. This done through the new sampler keyword in each of those procedures. This is a string specifying the sampler name (e.g. "A3"). The sampler name is synonymous with a single combination of ifnum, plnum, and fdnum. It is an error to use both types of data selection (sampler name and any of IF, polarization, or feed number) in the same call to one of these routines. Most users will continue to use the integer numbers to select the data to handle but in some cases users may find it more convenient to select data by sampler name.
|
>
> |
- Data selection by sampler name is now available in the following routines: flag, getbs, getfs, getnod, getps, getsigref and gettp. This is done through the new sampler keyword in each of those procedures. This is a string specifying the sampler name (e.g. "A3"). The sampler name is synonymous with a single combination of ifnum, plnum, and fdnum. It is an error to use both types of data selection (sampler name and any of IF, polarization, or feed number) in the same call to one of these routines. Most users will continue to use the integer numbers to select the data to handle but in some cases users may find it more convenient to select data by sampler name.
|
| Changed: |
<
< |
- Because the scan number may now be a larger, the format of the output of header as well as the header information on plots has been adjusted.
|
>
> |
- Because the scan number may now be larger, the format of the output of header as well as the header information on plots has been adjusted.
|
|
|
| Changed: |
<
< |
*
- New data file index. The index version number has changed because of the new data selection fields. When you open a data file with an index having an older version number GBTIDL will automatically recreate the index from the data file without any loss of information.
-
- New arguments to existing procedures.
|
>
> |
- Data selection by sampler name is now available in the following routines: flag, getbs, getfs, getnod, getps, getsigref and gettp. This done through the new sampler keyword in each of those procedures. This is a string specifying the sampler name (e.g. "A3"). The sampler name is synonymous with a single combination of ifnum, plnum, and fdnum. It is an error to use both types of data selection (sampler name and any of IF, polarization, or feed number) in the same call to one of these routines. Most users will continue to use the integer numbers to select the data to handle but in some cases users may find it more convenient to select data by sampler name.
- The structure returned by scan_info now includes N_SAMPLERS and SAMPLERS fields giving the number of unique samplers and the array of unique sampler names in that scan.
- Two new functions, samplerinfo and whichsampler can be used to convert between an (ifnum, plnum, fdnum) combination and sampler.
- As reported in the sdfits release notes several columns are now wider. The OBJECT column is now 32 characters (was 16 characters) and the SCAN, LASTON, LASTOFF, INT, and NSAVE integer columns are now long integers (32 bits) instead of short integers (16 bits). The INT and NSAVE columns are not written by sdfits but are written by GBTIDL (keep and nsave). NOTE if you are saving to an existing, older, file it will still have the short integer columns and consequently you can lose information if you use a large NSAVE value. In order to take advantage of these larger integers it is necessary to write to a new output file.
- New data file index. The index version number has changed because of the new long integer columns also found in the index (SCAN, INT, and NSAVE). GBTIDL will re-create the index file as necessary when an sdfits file is opened and the existing index file has an older version number. There will be an informational message when that happens.
- Because the scan number may now be a larger, the format of the output of header as well as the header information on plots has been adjusted.
- A procedure, oplotfn, allows you to over-plot a function on the GBTIDL plotter. You supply an IDL function name and the parameters used to generate the values to be plotted. The GBTIDL plotter then invokes this procedure using those parameters as needed to generate the values to be plotted. This allows for a smooth display of the function at any zoom level. The Gaussian plotting in GBTIDL has been modified to use this new procedure.
- New arguments to existing procedures
- unflag has a new /all keyword. When used, all flags are unflagged.
- bias has a new buffer keyword to identify which buffer to apply the bias to (the default is the primary data container, buffer 0).
|
| Changed: |
<
< |
*
|
>
> |
- molecule, recombc, recombhe, recombh*, recombn*, and recombo had a bug that resulted in the marked lines some times appearing at the wrong position. This would occur when the plotted x-axis included an additional offset (usually the source velocity). The shift due to that offset was applied in the wrong direction.
- avgstack with the /noclear keyword set was not behaving as advertized. It was correctly not clearing the accumulation at the start but incorrrectly clearing it at the end.
- In scale the buffer keyword could be unexpectedly set when the procedure was invoked. The buffer keyword is intended as an input keyword, meaning that the procedure itself should just use this value and not change it internally. In some cases, scale was changing the value of buffer to 0. This would happen when scale was intentionally called with an unset value for the buffer keyword. That can be a useful technique when used within another procedure which itself has a buffer value. That could cause unexpected results if that value of buffer was subsequently used or tested.
|
| Changed: |
<
< |
*
|
>
> |
- Use of the IDL plot procedure may some times cause the GBTIDL oplot procedure to fail. This will be fixed in the next release of GBTIDL.
getrec, 0
plot, x, y1
oplot, x, y2
* But in this example, oplot fails
plot, x, y1
getrec, 0
oplot, x, y2
|
| Changed: |
<
< |
- You must have an IDL license to run GBTIDL. IDL version 6.0 and later is supported. If you don't have a license you can run GBTIDL remotely, using the computers in CV or GB. See the GBTIDL homepage for details. IDL 6.3 is needed for large file support on a Mac.
|
>
> |
- You must have an IDL license to run GBTIDL. IDL version 6.0 and later is supported. If you don't have a license you can run GBTIDL remotely, using the computers in CV or GB. See the GBTIDL homepage for details. IDL 6.3 or later is needed for large file support on a Mac.
|