TABLE Instruction

TABLE( options ) start end listofseries

TABLE computes basic statistics on several series. The statistics computed are the number of observations, sample mean, sample standard error, minimum and maximum. It also computes the overall maximum and minimum values of the whole set of series. For a broader range of statistics on a single series, see STATISTICS.

Wizard

If you create a Series Window, you can also get a statistics table by selecting the desired series and clicking on the

Parameters

start, end	estimation range. By default, the maximum range of each series separately.
listofseries	list of series to include. If left blank, all current series.

Options

SMPL=Standard SMPL option [unused]

WEIGHT=Standard WEIGHT option [unused]

If you use either option, it applies to all series in the list.

[PRINT]/NOPRINT

TITLE=”title for output” [none]

Use NOPRINT if you want to suppress the displaying of the output on the screen. This will only make sense if you are TABLE to get the maximum and minimum across the list or are using the MATRIX option. Use TITLE to supply your own title to label the resulting output.

WINDOW="title of window"

If you use the WINDOW option, the output goes to a Report Window with the given title, rather than being inserted into the output window or file as text. Note that even without the WINDOW option, you can reload the report from the Reports Windows list on the Window menu.

PICTURE="picture code for statistics"

A picture code lets you control the formatting of the numbers in the output.

MATRIX=RECTANGULAR array for statistics [unused]

You can use the MATRIX option to save the computed statistics into a K × 5 array, where K is the number of variables in the listofseries. In each row, column one contains the number of observations, column 2 the mean, column 3 the standard error, column 4 the minimum value, and column 5 the maximum value.

Variables Defined

%MAXIMUM	maximum value found in the listofseries (REAL)
%MINIMUM	minimum value found in the listofseries (REAL)

Notes

It’s always a good idea to use a TABLE instruction immediately after the DATA instruction, particularly the first time you use a data set. It gives you a quick way to check whether your data are in good shape. For instance, you can easily detect series which have missing data because the number of observations will not match with those for other series. You might also have a series whose data range doesn’t match with what you expect. TABLE can help you spot series which have problems, but you will have to use PRINT or some other instruction to isolate the cause.

The sample standard error is computed using an N–1 divisor where N is the number of data points in a series.

%MAXIMUM and %MINIMUM are the only variables that TABLE defines that you can access from within your program (except by using the MATRIX option). Use the STATISTICS instruction applied to the series one at a time if you need access to such things as the mean and number of observations.

Examples

This uses TABLE for descriptive statistics on two subsamples of a data set.

* UNION.RPF

* Adapted from Johnston and DiNardo(1997), pp 415-425

open data cps88.asc

data(format=prn,org=columns) 1 1000 age exp2 grade ind1 married $

lnwage occ1 partt potexp union weight high

table(smpl=union,title="Statistics for Union Members") / $

potexp exp2 grade married high

table(smpl=.not.union,title="Statistics for Non-Union Members") / $

potexp exp2 grade married high

This uses TABLE to get %MAXIMUM and %MINIMUM for the set of series so that all graphs can be done on the same scale.

* SPGRAPH.RPF

cal(m) 1996

open data oecdsample.rat

data(format=rats) 1996:1 1998:12 canusxsr frausxsr jpnusxsr gbrusxsr

* Flip to dollars per foreign unit and assign descriptive labels

set canusxsr = 1.0/canusxsr

set frausxsr = 1.0/frausxsr

set jpnusxsr = 1.0/jpnusxsr

labels canusxsr frausxsr jpnusxsr gbrusxsr

# "Canada" "France" "Japan" "UK"

* Convert to percent appreciation

dofor i = canusxsr frausxsr jpnusxsr gbrusxsr

compute base=([series]i)(1996:1)

set i = 100*(i{0}/base - 1.0)

end dofor i

* Use TABLE to get the maximum and minimum across all series

table(noprint) / canusxsr frausxsr jpnusxsr gbrusxsr

* Set up the SPGRAPH for a 2x2 matrix

spgraph(vfields=2,hfields=2,header="U.S. Dollar vs Major Currencies",$)

subheader="Percent appreciation over the sample")

dofor i = canusxsr frausxsr jpnusxsr gbrusxsr

graph(max=%maximum,min=%minimum,header=%l(i)) 1

# i

end dofor

spgraph(done)

Sample Output

These are from the first example.

Statistics for Union Members

Series Obs Mean Std Error Minimum Maximum

POTEXP 216 22.77777778 11.41710942 1.00000000 49.00000000

EXP2 216 648.57407407 563.34258574 1.00000000 2401.00000000

GRADE 216 12.56018519 2.27342339 5.00000000 18.00000000

MARRIED 216 0.75000000 0.43401854 0.00000000 1.00000000

HIGH 216 0.76388889 0.42567775 0.00000000 1.00000000

Statistics for Non-Union Members

Series Obs Mean Std Error Minimum Maximum

POTEXP 784 17.81122449 12.93830834 1.00000000 55.00000000

EXP2 784 484.42602041 595.76310328 1.00000000 3025.00000000

GRADE 784 13.13903061 2.67619124 0.00000000 18.00000000

MARRIED 784 0.61096939 0.48784152 0.00000000 1.00000000

HIGH 784 0.51403061 0.50012216 0.00000000 1.00000000