Workbook for SLD Offline Users - Data Exercise 2: Exploring Datacats

All official public SLD Datacats are stored in the directory DUCSROOT:[SLD.DATACAT] or if you want to use all logicals DUCSROOT:DUCSDATACAT. So lets go there and see what we can find.

From the VMS prompt type:

   > Set Default DUCSROOT:DUCSDATACAT
   > DIR

We can tell the type of events in a Datacat by the first few letters in its name. For instance:

ACQ = Acquisition Data (raw data from the detector not to be used by regular users)
BHA = Bhabba events
DISKMDST = Reconstructed Mini-Dst events on disk (this is the form of data most commonly used for Physics analysis)
HAD = Raw events that passed the SLD Physics filter
MC = Monte Carlo events
MUP = Mu Pair events
RAN = Random events or cyclic trigger events
RAW = Raw events (an exact copy of ACQ data, RAW is meant for general use)
RECON = Reconstructed events that passed the Physics filter, stored in DST format (these events are missing the constants banks and so are not "Physics Ready")
TRK = Tracking information
WAB = Wide Angle Bhabba events

There are more Datacats than those listed above, however these are most of the basic ones. The numbers following the name can refer to the version of the reconstruction and the year the sample is from.

Reading a Datacat

Within a Datacat we can get information about Run numbers, the number of events, the name of the file its stored in, the name of the tape it is stored on and comments about the file. In order to fully read a datacat one needs a wider terminal then the standard 48 rows by 80 columns. In order to enlarge your window issue the VMS command:

   > PAGESIZE 48 110

Now that your window is large enough lets look at the Datacat corresponding to the raw 1993 Data that passed the physics filter. In order to do that type:

   > EVE DUCSDATACAT:HAD93.DATACAT

You should see something that looks like:


   ! SECTION: DATACAT
   ! 1993 Polarized Run FILTER files -- HAD stream
   !==============|======|===========|=======|===========|==========|=========|========|==========|=============
   !Dataset       |Medium| Tape Name |File # | Tape label| File     | Status  |   #    |  Entry   | Comments
   !Name          |      | Disk Name |Diskadd| File name | Type     |         | Events |  Date    |
   !------------------------------------------------------------------------------------------------------------
   |HAD20766      | CART | QW0344    |     1 | HAD20766  |          |   OK    |    124 |  3/31/93 |
   |HAD20768      | CART | QW0344    |     2 | HAD20768  |          |   OK    |    106 |  4/01/93 |
   |HAD20777      | CART | QW0344    |     3 | HAD20777  |          |   OK    |     70 |  4/02/93 |
   |HAD20778      | CART | QW0344    |     4 | HAD20778  |          |   OK    |    181 |  4/01/93 |
   |HAD20780      | CART | QW0344    |     5 | HAD20780  |          |   OK    |     54 |  4/02/93 |
   |HAD20793      | CART | QW0344    |     6 | HAD20793  |          |   OK    |    143 |  4/02/93 |

The top lines give you a brief description of the type of events in the files in the datacat. In this case polarized 93 runs that passed the filter. The next set of lines gives you the column names. The following rows are the data.

Dataset Name

The Dataset name or "Nickname" is an alias given to a file. It is possible to open a tape by specifying the Dataset name only. For instance OPENTAPE READ HAD20766 STAGE WAIT would open the first tape in the dataset. One could also use OPENTAPE READ QW0344.1/HAD20766 CART STAGE WAIT to open the same file. The Dataset name is arbitrary (although its nice to use the run number in the name) and not necessarily unique. For instance one could name every file in their datacat "BOB". Then the command OPENTAPE READ BOB STAGE WAIT would sequentially read all files with the dataset name "BOB". This is actually a good trick to remember for when you have a number of tapes with similar types of data. Another way to open multiple tapes using the Dataset name would be to issue the command OPENTAPE READ HAD[20766-20780] STAGE WAIT. That would open all tapes with Dataset names between HAD20766 and HAD20780.

Medium

This column tells you whether the file is stored on disk or tape/cartridge.

Tape Name Disk Name

The name of the tape or disk that the data is stored on. Remember both a tape and a Disk can store multiple files so this name alone will not suffice to open either.

File #

The Number of the file on tape. In order to specify the first file on tape QW0344 we would type QW0344.1 . So the file number is specified after the Tape name and a period.

Tape Label or Filename

This is the official name of the file to be opened. It is just a fluke of this datacat that it matches the Dataset name, this is not always the case. So the official name of the first file is QW0344.1/had20766.

Status

There must be an OK in this column or you will be unable to open this tape.

Number of Events and Entry Date

This should be self explanatory!

Comments

In this case the comments column is empty, but it is often useful to make use of it especially if there is something strange about the file.

Making Your Own Private Datacat

The material in a particular Datacat is not mutually exclusive a particular file may appear in a number of Datacats. Furthermore there is nothing magical about the official datacats stored in DUCSROOT:DUCSDATACAT. Users are allowed to have there own private version of any of these datacats.

In fact it is most often desirable to have private datacats because IDA will always search through the users account for datacats and entries before going to DUCSDATACAT. Since there are many Datacats to search through in DUCSDATACAT files are usually found and opened faster if they appear in private datacats.

For now we have no private tapes yet but we will get some later in this workbook so let us get a jump on things and get our own Datacat. When we issued the DIR command you might have noticed an entry:

   NEWUSERS.DATACAT;1         1  22-DEC-1994 13:35:21.84

This datacat contains the entry for the data file we have and will be using. At this time we may want to copy it to our own directories. To do this SET DEFAULT $USR:[YOURDIRECTORY]. Once you are back in your own directory type:

   > COPY DUCSDATACAT:NEWUSERS.DATACAT MY.DATACAT

Feel free to look at this file in the Editor it should look something like:

   ! SECTION: DATACAT
   !
   !    This DATACAT references data used in the Offline Users Workbook
   !===========|=========|=================================|=======|===============|=====|======|========|==========|======
   !Dataset    | Medium  |           Tape Name             |File # |  Tape label   |File |Status|   #    |  Entry   | Comme
   !Name       |         |           Disk Name             |Diskadd|  File name    |Type |      | Events |  Date    |
   !---------------------------------------------------------------------------------------------------------------------
   |REC94_MDST | VMSDISK | DISK$SLD_FAC0:[SLDWWW.WORKBOOK] |       | REC94V11_MDST |     | OK   | 500    | 1/9/94   |

When the time comes and you have your own tape make up a dataset name and enter it in this datacat. You may also want to enter files you find yourself frequently using.

Back to Main Data Page

Eric Weiss
13 January 1995