INTEGER FREQUENCY TABLE

INTEGER FREQUENCY TABLE (LET)

Let Subcommand

Create a frequency table for integer values.

This command is intended for binning data for discrete distributions. In partiucular, this command is typically used before performing a chi-square goodness of fit.

There are two primary features for the binning:

1. Discrete distributions typically expect the data to occur at integer values. This command will bin the data at the integer values in the data.

2. The usual recommendation for the chi-square goodness of fit test is that the minimum class frequency should be at least five. This command will automatically combine bins with small frequencies. The default minimum class frequency is five, although this can be set to a different value.

LET <y2> <xlow> <xhigh> = INTEGER FREQUENCY TABLE <y>
                        <SUBSET/EXCEPT/FOR qualification>
where <y> is a variable containing the raw data;
            <y2> is a variable where the class frequencies are stored;
            <xlow> is a variable where the lower limits for the class bins are stored;
            <xhigh> is a variable where the upper limits for the combined bins are stored;
and where the <SUBSET/EXCEPT/FOR qualification> is optional.

LET Y2 XLOW XHIGH = INTEGER FREQUENCY TABLE Y

You can specify the minimum frequency for a bin by entering the command
LET MINSIZE = <value>

before the INTEGER FREQUENCY TABLE command.

The following commands currently accept grouped data with either equal size bins or unequal size bins:
PROBABILITY PLOT
PPCC PLOT
KS PLOT (plots minimum chi-square statistic for discrete distributions)
CHI-SQUARE GOODNESS OF FIT

The minimum frequency for a bin defaults to 5.

None

BINNED	= Convert raw data to grouped data.
COMBINE FREQUENCY TABLE	= Combine clases with small frequencies in frequency tables with equal sized bins.
CHI SQUARE GOODNESS OF FIT	= Perform a chi-square goodness of fit test.
PROBABILITY PLOT	= Generate a probability plot.
PPCC PLOT	= Generate a ppcc plot.
KS PLOT	= Generate a Kolmogorov-Smirnov (or chi-square) goodness of fit plot.

Distributional Modeling

2006/5

 
LET K = 3
LET LAMBDA = 0.9
LET Y = BOREL TANNER RANDOM NUMBERS FOR I = 1 1 500
LET Y2 XLOW XHIGH = INTEGER FREQUENCY TABLE Y
LET LAMBDA1 = 0.5
LET LAMBDA2 = 0.95
Y1LABEL CHI-SQUARE
X1LABEL LAMBDA
BOREL TANNER KS PLOT Y2 XLOW XHIGH
JUSTIFICATION CENTER
MOVE 50 5
TEXT LAMBDAHAT = ^shape, Minimum Chi-Square = ^minks