PACKAGE |STAT Data Manipulation and Analysis, by Gary Perlman
NAME features - tabulate features of items
USAGE features [-h] [-c col-delim] [-f inputfile] [-i itemwidth] [-l leader] [-n no-str] [-p pagesize] [-w featwidth] [-y yes-str] feature[=width] [feature[=width] ...]
WEB FORM features can be run from an online web form.
DESCRIPTION Think of this program as a consumer product comparison checklist tool. For each item, there are features that you would like the product to have, and for some features, several values are possible.
OPTIONS
feature
Feature Names. Feature names indicate which features to tabulate, and in which order, to describe items. Features on the command line can be exact matches for item's features, but to simplify the format of feature files, features in files can be abbreviated and upper and lower case letters are treated as equivalent. For example, Usability on the command line matches usab=good in the feature file. You can specify the width of a column for a feature by putting a width after an equal sign. See the examples.
-c col-delim
Column Delimiter String. Default: bar (|). This string is placed between every feature column describing an item. It can be wider than one character; for example, " | " would give a wider table with more space in columns.
-f inputfile
Input file. Default: standard input. This option allows you to specify a file name for the item description lines.
-h
Use Horizontal Format for Headers. If this option is selected, then all field widths for features must be specified.
-i itemwidth
Width of Items. Default: 15 characters wide. If you have ones that are wider, the tables won't line up correctly. This option allows you to widen the space used to show the items.
-l leader
Leader Character After Item Name. Default: underbar (_).
-n no-str
String to Print When Item Feature is Not Present. Default: space.
-p pagesize
Size of Pages. Default: 60 lines. For every page, a new feature-table header is generated.
-w feat-width
Width of Feature Columns. Default: 1 character. This option will allow you to set all the feature column widths. Individual column widths can be overridden with feature widths on the command line.
-y yes-string
String to Print When Item Has a Feature. Default: sharp (#). This can be changed to any character. It can also be changed for an individual item-feature pair by using a feature=value format. For example,
Perlman eyes=H
could be used to indicate that Perlman's eyes are hazel. For longer strings, you will want to adjust the width of the columns with the -w option or with the feature name on the command line.

The following standard help options are supported. The program exits after displaying the help.
-L
Display limits
-O
Display options and values
-V
Display version number and date
EXAMPLES Product Comparisons

Suppose you want to buy a toaster with a bunch of features like inexpensive, reliable, wood-grain and one that holds four slices of bread. These features might be: price (L=low, M=medium, H=high), reliability (yes, no), finish (M=metal, W=wood-grain, O=other), and capacity (number of slices of bread). A set of toasters (items) might be described by a list of features:

BrandX    price=Med reliable finish=Wood capacity=4
BrandY    price=High finish=Metal capacity=2
BrandZ    price=Low finish=Wood capacity=4
The above input, along with the feature labels:
Price Reliable Finish Capacity
in the command:
features -c " | " Price=4 Reliable Finish=6 Capacity
will display:
                | Price
                |      | Reliable
                |      |   | Finish
                |      |   |        | Capacity
BrandX_________ | Med  | # | Wood   | 4 |
BrandY_________ | High |   | Metal  | 2 |
BrandZ_________ | Low  |   | Wood   | 4 |

Another command shows how the same information can be viewed differently.

features -l. -c" " -h -nNo -yYes Price=6 Reliable=10 Finish=6 Capacity=10
                Price  Reliable   Finish Capacity
BrandX......... Med    Yes        Wood   4
BrandY......... High   No         Metal  2
BrandZ......... Low    No         Wood   4
Other Uses

The program has been useful for organizing lists of publications by the topics they cover and for listing the research interests of faculty in a large Midwest computer science department.

SEE ALSO ptx (permuted index), sort
WARNINGS Try to avoid similar feature names, especially ones that begin with the same characters.
UPDATED March 10, 1989