DataMuseum.dk

top - metrics - download

    Length: 3709 (0xe7d)
    Types: TextFile
    Names: »desc«

└─⟦87ddcff64⟧ Bits:30001253 CPHDIST85 Tape, 1985 Autumn Conference Copenhagen
    └─⟦this⟧ »cph85dist/stat/doc/cat/desc« 

TextFile

DESC(1)                UNIX User's Manual                 DESC(1)

NAME
     desc - describing a single distribution

SYNOPSIS
     desc [-i N  -m N  -M N  -cfph -o -s -F N]

DESCRIPTION
     _▶08◀d_▶08◀e_▶08◀s_▶08◀c does a number of summary statistics and frequency
     counts.  _▶08◀d_▶08◀e_▶08◀s_▶08◀c reads from the standard input (via redirection
     with < or piped with |) and writes to the standard output.
     The program's input is a series of numbers separated by new-
     lines, tabs or spaces.  Summary statistics can be obtained
     without use of program arguments.

OPTIONS
     The options available with _▶08◀d_▶08◀e_▶08◀s_▶08◀c allow you to request various
     types of descriptions of your data, as well as control the
     format of these descriptions.  In the following section,
     when an option is followed by N, a numerical argument is
     expected.  Program arguments (options) can be combined in
     any number or order.

     -c   changes f and p options to CUMULATIVE.

     -f   table of midpoints and FREQUENCIES.

     -F N requests a F-TEST of Ho: mean = N (also t).

     -h   prints a HISTOGRAM with frequencies.

     -i N sets the INTERVAL WIDTH for h, f & p.

     -m N sets the MINIMUM number of the first interval.  All
          numbers below N are IGNORED ON INPUT.

     -M N sets the MAXIMUM number allowed to N.  Numbers larger
          than N are IGNORED ON INPUT.

     -o   requests ORDER statistics (also S).

     -p   table of midpoints and PROPORTIONS.

     -s   summary STATISTICS according to SPSS formulas.

     -S   requests order STATISTICS

     -v   output statistics in VARIABLE (name=value) format (use-
          ful for input to CALC calculator).

     When computing frequencies, numbers on bin boundaries end up
     in the next lower bin, except for the numbers on the lower
     boundary of the first bin, which are put in the first bin.

Printed 5/30/85           March 5, 1985                         1

DESC(1)                UNIX User's Manual                 DESC(1)

     Thus, the first bin is a doubly closed interval, while all
     the others are half open: [],(],(],(],(],...(]

LIMITATIONS and DIAGNOSTICS
     _▶08◀d_▶08◀e_▶08◀s_▶08◀c does its computations in either one or two passes,
     depending on whether the data is stored.  If the -i and -m
     options are used, then it is not necessary to store the data
     for any frequency table.  If not, or if you want order
     statistics (which requires storing and sorting the data),
     then the number of input points is limited to some large
     number, such as 10,000.  The number of bins is limited to
     some large number, such as 1000.

EXAMPLES
                             desc < file
     prints basic statistics (same as desc -o < file)

                      desc -cf -m 0 -i 10 < file
     prints a table of midpoints and CUMULATIVE FREQUENCIES.  The
     MINIMUM allowable number is 0, and the INTERVAL WIDTH is 10,
     so the first interval is [0,10] and the next is (10,20], and
     so on.  Intervals will be added to accommodate the largest
     numbers.  Because i and m are specified, there is no limit
     on the number of points.

            desc -s -F 0 -m 100 -M 1000 -i 100 -cp < file
     first prints summary STATISTICS.  Then it prints a F-VALUE,
     and prob (F).  Finally, desc will print a table of interval
     midpoints and CUMULATIVE PROPORTIONS.  The MINIMUM allowable
     number is 100, the MAXIMUM allowable number is 1000, and the
     INTERVAL WIDTH is 100.

SEE ALSO
     unixstat(1), dm(1)

AUTHOR
     Gary Perlman

KEYWORDS
     descriptive statistics, summary statistics, data analysis

Printed 5/30/85           March 5, 1985                         2