A QUICK INTRODUCTION TO PAW

Introduction

PAW (Physics Analysis Workstation) is a CERN application used to analyze and plot data. There is a full writeup on PAW available from CERN (see the "PAW" link above), as well as a "help" command in PAW, but you generally have to have an idea of the command you're looking for before the official documentation is of much use.

Also, there is a very comprehensive PAW home page at CERN which contains almost all of the information found here, and much, much more.

The example kumac file given below will introduce you to some common PAW commands.

Note: To run paw, you need to "setup cern", and execute "pawX11".

Command Structure

PAW commands are organized in a tree structure. For example

tgroup/sgroup/command {arguments}

means execute the command command in the subgroup sgroup of the top group tgroup. If command is unique among all groups and subgroups, it is not necessary to specify the group:

command {arguments}.

Data Structures

There are several means of organizing and manipulating data in PAW:

• Histograms - A series of "bins," each with an associated value. One of the most common uses of histograms is to represent the distribution of data in a specific variable or variables. Histograms can be either 1 or 2 dimensional, but only 1-D histograms can have have uncertainties (i.e., error bars) stored with the value of each bin.
• Vectors - A series of numbers with up to 3 indices - corresponds to "arrays" in languages like C and FORTRAN.
• Ntuples - A series of numbers, grouped into data points. See the CERN HBOOK writeup for information about Ntuples.
• Functions - 1 or 2 dimensional functions.

Kumacs and Macros

Every PAW command can be executed in a command file called a KUMAC. (Kuip MACROS). See this example, which is also a very basic PAW tutorial.

An important macro is .pawlogon.kumac in your home directory, which executes your favorite commands each time you start a PAW session. Another useful macro is last.kumac, which is created by PAW, and contains a record of the commands executed in each session.

Please read the MACRO section of the online PAW help and PAW documentation.

Options and Attributes

See the help for the option and the set commands for information on setting attributes such as the size and font of histogram titles, whether to use logarithmic or linear axes, etc. There is a very useful graphical display of these attributes in the PAW FAQ and the PAW documentation.

Hard Copies

To make a hard copy, do the following:

fort/file lun {filename.ps} - open a file on logical unit lun. (The logical unit is similar to a file descriptor, but you choose it. It's a number from 1 to 99.) The filename doesn't have to end in .ps, but it's a good idea since it will be a postscript file.

meta lun - make the file you just opened as a postscript meta file.

Make your plots.

close lun - close the meta file.

You are now ready to view the file with ghostview or send it to a printer.