Home of the original IBM PC emulator for browsers.
[PCjs Machine "ibm5170"]
Waiting for machine "ibm5170" to load....
A statistical analysis package for handling numerical data, operated by
entering one-line commands and subcommands. Command ``batch'' files can
be created for automatic execution, along with explanatory screen
remarks.
STATMATE operates on information contained in a database, generated by
the program. A user ID is required before entering a database, and for
every new user ID, an empty database is created. This feature permits
multiple users to work with STATMATE while keeping the data files
separated.
Extract data from an ASCII text file and load it into the database for
operation. Data is stored in columns and rows, and you can extract
portions of the data according to your specifications. As you
manipulate the data, the results can be displayed on the screen,
printed, or saved on a disk file.
The main analytic features are elementary statistics, scatter plots,
cross tabulations, histograms, data comparison using the T-Test,
correlation, arithmetic operations, distribution functions, curvilinear
regression, multiple regression, nonlinear regression, data recoding,
and data transformation and manipulation. An on-line help facility is
included to give you a detailed description of all the STATMATE
commands.
Disk No: 861
Program Title: STATMATE/PLUS version 1.3 (Disk 1 of 3)
PC-SIG version: 1.4
STATMATE is a three-disk statistical analysis package for handling
numerical data, and is operated by entering one-line commands and
subcommands. Command "batch" files can be created for automatic
execution, along with screen remarks to explain the data manipulation on
the terminal.
The data STATMATE operates on is contained in a database, which is
generated by the program. A user ID is required before entering a
database, and for every new user ID, an empty database is created. This
feature permits multiple users to work with STATMATE while keeping the
data files separated.
The size of each database can be changed with the installation program.
STATMATE lets you extract data from an ASCII text file and load it into
the database for operation. Data is stored in columns and rows, and you
can extract portions of the data according to your specifications. As
you manipulate the data, you can have the output on results displayed on
the screen, printed, or saved on a disk file.
The main analytic features available with the STATMATE package are
elementary statistics, scatter plots, cross tabulations, histograms,
data comparison using the T-Test, correlation, arithmetic operations,
distribution functions, curvilinear regression, multiple regression,
nonlinear regression, data recoding, and data transformation and
manipulation. An on-line help facility is included to give you a
detailed description of all the STATMATE commands.
You will need disks #861, #862 and #863 to run this program.
Usage: Statistics Analysis.
Special Requirements: Two disk drives.
How to Start: Type GO (press enter).
Suggested Registration: $50.00
File Descriptions:
SMATE EXE Main program.
SMINSTLL EXE Installation program.
SMQDATA COM Data preparation program.
SMHELP TXT Help file.
DEMO Example command file.
README How to get started.
SMSA$ Configuration file.
??????? DAT Sample data file (6 files).
??????? OVR Overlay file (17 files).
PC-SIG
1030D E Duane Avenue
Sunnyvale Ca. 94086
(408) 730-9291
(c) Copyright 1987,88,89 PC-SIG, Inc.
╔═════════════════════════════════════════════════════════════════════════╗
║ <<<< Disk #861 STATMATE/PLUS (Disk 1 of 3) >>>> ║
╠═════════════════════════════════════════════════════════════════════════╣
║ For documentation refer to disk number 863. ║
║ ║
║ To start STATMATE/PLUS, Type: SMINSTLL (press enter) ║
╚═════════════════════════════════════════════════════════════════════════╝
HELP
---COMMANDS---
BREAKDOWN COMPUTE CHART CORRELATE CROSSTAB CURVE
CUSUM EDIT ELSE END ERASE EXECUTE
EXIT GIVE HELP HISTOGRAM INPUT KOLMOGOROV
LET NONLINEAR ONEWAY ONPARAM PLOT POLYNOMIAL
PRINT QUERY RCORRELATE RECODE REGRESSION REMARK
SET SHOW STATISTICS STEPWISE TNPARAM TTEST
TWOWAY WHEN WRITE
Hints:
1. Enter HELP QUERY, for example, to get help on QUERY.
2. Ordinarily STATMATE is terminated with the EXIT command.
However, it may be also by terminated by a control-c character.
3. A command line may be continued onto another line by
ending a line with a comma.
4. In the command descriptions, parenthesises denote the
default option for a modifier. For example,
TYPE : (FITTED), RESIDUAL, FORECAST
indicates FITTED is the default for TYPE.
5. The variable #0 is predefined in STATMATE and represents
the variable containing 1,2,3,4,...
COM
COMPUTE
Purpose: Computes residuals, fitted, and forecast values after
completion of REGRESSION, CURVE, or POLY
Examples:
COMPUTE XRESLT,TYPE=RESIDUAL
COMPUTE SALESCAST,TYPE=FORECAST
Modifiers:
TYPE : (FITTED), RESIDUAL, FORECAST
Hints:
1. The REGRESSION, CURVE, or POLY commands must be used to
first calculate the coefficients which are used by COMPUTE.
2. COMPUTE remembers the set of independent variables (X) and
the dependent variable (Y) used to compute the coefficients
in the regression analysis last used.
3. Missing values in the COMPUTE variable and the dependent
variable influence whether the resulting value is a missing
value or a computed value.
4. Does not apply to NONLINEAR command's models.
CORR
CORRELATE
Purpose: Computes pair-wise correlations
Examples:
CORRELATE URBANPOP,RURALPOP,...
CORRELATE DOSAGE,TIME,ORDER=YES
Modifiers:
SIGNIFICANCE : (NO) , YES
ORDER : (NO) , YES
CROSS
CROSS TABULATION
Purpose: Produces a two-way crosstabulation
Examples:
CROSSTAB FACTORA ON FACTORB
CURVE
CURVE
Purpose: Computes curvilinear regression on two variables
Examples:
CURVE Y ON X,TABLE=FIT,ANOVA,PARAM,BEST=TRES
CURVE Y ON X,EQUATION=LIN,ORI,QUA,LOG
Modifiers:
EQUATION : (LIN), HY1, HY2, EXP, HY3, LOG, ORI, POW, QUA, SQU
ALL
TABLE : (FIT), (PARAM), ANOVA
ALL
BEST : (RSQ), TRES, ORES
EDIT
EDIT
Purpose: Allows data to be edited
Examples:
EDIT YEAR,RESPONSE
Hints:
1. Only the variables following EDIT may be edited with
the EDIT subcommands.
END
END
Purpose: Ends a WHEN command condition -- use HELP WHEN
Examples:
END
ELSE
ELSE
Purpose: Reverses a WHEN command condition -- use HELP ELSE
Examples:
ELSE
ERASE
ERASE
Purpose: Erases variables from the database
Examples:
ERASE SALES
ERASE #12 THRU END
Hints:
1. Either a single variable or a set of consecutive variables
can be erased. For example, ERASE A THRU C is permitted
but ERASE A,C is not.
EXE
EXECUTE
Purpose: Allows "batch" execution of commands
Examples:
EXECUTE CMDFILE
Hints:
1. The file named must be an ASCII file.
EXIT
EXIT
Purpose: Exits from STATMATE to the operating system
Examples:
EXIT
GIVE
GIVE
Purpose: Gives attributes (name, type, missing value) to a variable
Examples:
GIVE NAME #15,VAR3,VAR4 -- assign the names VAR3, VAR4 to #15 and #16
GIVE MISSING AGE,-1.0 -- interpret -1 as a missing value for AGE
GIVE TYPE #10,A,N,N,A -- make #10 numeric type, #11 alpha type, etc.
Hints:
1. To give either a "missing value" or type attribute to a variable
the the variable must first be erased (see ERASE).
HIST
HISTOGRAM
Purpose: Produces a histogram of a variable
Examples:
HISTOGRAM SALES,BARS=8,VPOS=15,RANGE=17 THRU 20
Modifiers:
RANGE : min THRU max
VPOSITION : number-of-lines (10)
TITLE : title
BARS : number of bars (10)
Hints:
1. Histogram divisions are determined by RANGE and BARS
2. Use single quotes (') around title
INPUT
INPUT
Purpose: Provides a way of entering data from a file
Examples:
INPUT MYFILE KEEP 2,OMIT 1,KEEP 1
INPUT DATAFILE.DAT KEEP 3 AT #22
Hints:
1. When reading in alpha data, make sure that you have used the
GIVE TYPE command for any variables which are to contain
alpha data.
2. Make sure the input file is an ASCII file with DATA as the
very first word on the first line, and no spaces before DATA.
3. A carriage return must appear at the end of the last line
of your data file.
4. OMIT and KEEP define which, and how many, fields are either
omitted or kept from the input file. Enough KEEP and OMIT
items must be specified to account for all fields in a record.
LET
LET
Purpose: Performs arithmetic operations, functions, or generates a
series of numbers and stores the result in a new variable
Examples:
LET NEWA=A+33.3
LET VOLUME=AREA*HEIGHT
LET XSQUARE=X**2
LET CONTROL=X-44.0
Hints:
1. Use HELP FUN for help on LET functions.
2. Complex arithmetic statements are not allowed. An expression
is composed of at most one arithmetic operator (* + - / **).
3. If a constant is involved it must appear after the operator.
For example, A+3.4 not 3.4+A.
4. The reciprocal of X can be obtained by using X**-1 since
1/X is not allowed.
FUN
LET FUNCTIONS
LET YEARLAG=LAG(YEAR,3) -- lag 3 periods
LET MYSAMPLE=SAM(STATES,12) -- randomly sample 12 items
LET SMOOTHED=MOV(SALES,4) -- moving average with 4 terms
LET TRANSFX=SQRT(X) -- square root
LET GROWTH=LOG(POPULATON) -- log base e
LET SHRINK=LOG10(STEMSIZE) -- log base 10
LET WGTINC=EXP(WGT) -- exponential
LET PERIOD=PRD(SALES,1,1,12) -- 1,2,3,4,5,6,7,8,9,10,11,12,1,2,...
LET ROWNUM=STP(RESP,4,1,2) -- 4,4,5,5,6,6,... (4=start,1=jump,2=prd)
LET ODDNUMBERS=NUM(MONTH,1,2) -- 1,3,5,7,... (1=start, 2=step)
LET NORMOUT=NOR(DOSE,5.0,1.4) -- Normal random nos. (mean=5, sd=1.4)
LET UNIFNUM=UNI(COST,20,44.5) -- Uniform random nos.(left=20,rght=44.5)
LET RUNDOWN=DNOR(LOSS,-40.2,4) -- Normal distribution points
LET DELAY=DEXP(DOSAGE,3.4) -- Exponential distribution points
LET WEAR=DWEI(MACHINE,2.2,4.1) -- Weibull distribution points
LET FIXLOSS=DCAU(CONT,2.4,8) -- Cauchy distribution points
LET ABSOLUTE=ABS(X) -- Absolute value
LET SUMIT=CUM(COST) -- Cumulative
LET WHOLE=INT(WGT) -- Integer or whole number
Hints:
1. The first argument of any function is a variable name.
ONEWAY
ONEWAY
Purpose: Performs a one-way ANOVA
Examples:
ONEWAY DOSAGE ON TREATMENT,METHOD=LSD,ALPHA=.05
Modifiers:
METHOD : (NONE), LSD, DUNCAN
ALPHA : (0.05)
POLY
POLYNOMIAL
Purpose: Perform polynomial regression
Examples:
POLY SPEED ON #0,TABLE=ALL
Modifiers:
TABLE : (PARAM), (FIT), ANOVA
ALL
DEGREE : (2)
PRINT
PRINT
Purpose: List variables
PRINT VAR1,VARY,SALES
PRINT #2,DOSAGE
Hints:
1. Prints any variable in the database
2. See SET command for hardcopy output
3. Use of the #0 variable is particularly useful here to
associate a case number with the data.
QUERY
QUERY
Purpose: Display database information on variables
Examples:
QUERY #3,#4,INCOME
QUERY #5 THRU #15
Hints:
1. No more than 20 variables may be queried with QUERY
RECODE
RECODE
Purpose: Changes the values of data to new values
Examples:
RECODE AGE,1 THRU 12 AS 1
RECODE MONTH,'1' THRU '3' AS 'Q1'
Hints:
1. Recodes numeric or alphanumeric data. Useful with
CROSSTABS.
REGRESS
REGRESSION
Purpose: Performs multiple regression on several variables
Examples:
REGRESSION Y ON X1,X2,TABLE=FIT,PAR,ANOVA,INT=YES
Modifiers:
TABLE : (PARAM), (FIT), ANOVA
ALL
INTERCEPT: (YES), NO
DURBIN : (NO), YES
REM
REMARK
Purpose: Provides a way to document a STATMATE session
Examples:
REMARK YOUR WISH IS MY COMMAND
SET
SET
Purpose: Sets a number of STATMATE parameters
Examples:
SET COPY=HARDCOPY
SET COPY=FILE MYRESFIL.OUT
SET COPY=OFF
SET OUT=COSTRESID
SET SEED=1532
Hints:
1. The HARDCOPY option sends output to your printer but not
to the terminal.
2. The FILE option sends output to the file but not to the
terminal.
3. The OUT option is used with NONLINEAR.
4. A seed value cannot exceed 30000.
5. A change in the seed value will cause the LET random number
functions to generate different sequences of numbers.
STAT
STATISTICS
Purpose: Performs simple statistical calculations
Examples:
STATISTICS #1,TABLE=SIMPLE
Modifiers:
TABLE : ALL or (SIMPLE)
TTES
TTEST
Purpose: Performs a T-test
Examples:
TTEST WEIGHT ON SEX
Hints:
1. The class variable (second variable) must contain only two classes.
TWOWAY
TWOWAY
Purpose: Performs a two-way ANOVA
Examples:
TWOWAY WGTGAIN ON FACTA,FACTB,DESIGN=NOREPEAT
Modifiers:
DESIGN : (NOREPEAT), REPEAT, NESTED, UNBALANCED
WHEN
WHEN
Purpose: Allows a subset of data for a variable to be selected
for analysis
Examples:
WHEN #0=50 THRU 100
WHEN AGE>10 AND SEX='MALE'
WHEN AGE=10 THRU 40
WHEN YEAR=1982,1984,1986
Hints:
1. Only simple expressions are allowed, i.e., one AND or OR.
For example,
AGE>10 AND WGT=100 THRU AND SEX='MALE' is not allowed.
2. Two successive WHENs are not allowed without an intervening END.
3. The END command terminates the condition established by a WHEN.
4. The ELSE command reverses the condition established by a WHEN.
WRITE
WRITE
Purpose: Writes data to a file
Examples:
WRITE YIELD,BONDS,STOCK ON MYSTOCK
Hints:
1. The output is written as an ASCII file which can be read
back into STATMATE with the INPUT command.
2. By removing the first line of output to the file with an
editor, the file can be used to input the data into other
programs. For example, dBASEIII.
TNPAR
TNPARAM
Purpose: Performs a two-way nonparametric ANOVA
Examples:
TNPAR RESPONSE ON GROUP1,GROUP2,TEST=COCHRAN
Modifiers:
TEST : (FRIEDMAN), COCHRAN
Hints:
1. In Cochran test, all values must be zero or one.
ONPAR
ONPARAM
Purpose: Performs a one-way nonparametric ANOVA
Examples:
ONPARAM XYZ ON LABCLASS
Modifiers:
none
KOLMO
KOLMOGOROV
Purpose: Performs a Kolmogorov-Smirnov one- or two- sample test
Examples:
KOLMO AFACTOR,DIST=WEIBULL,UPARAM=1.4,SPARAM=2.0
KOMLO AFACTOR,BFACTOR
Modifiers:
DIST : (NORMAL), UNIFORM, WEIBULL, CAUCHY, EXPONENTIAL
UPARAM : (0.0)
SPARAM : (1.0)
Hints:
1. normal: u is mean, s is std. dev.
2. exponential: u = displacement, s = std. dev.
3. Cauchy: u=median, u-s=1st quartile, s>0
4. uniform: u=lft end, s=rgt end
5. Weibull: u=shape, s=scale
6. Use DIST, UPARAM, and SPARAM and only one variable for
the one-sample test.
7. DIST, UPARAM and SPARAM do not apply to the two-sample test.
PLOT
PLOT
Purpose: Produces scatter diagrams
Examples:
PLOT Y1,Y2... ON X,TITLE='DOSAGE VS YEAR'
PLOT SALES ON YEAR,HRANGE=1900 THRU 1990
PLOT ?
Modifiers:
HRANGE : min THRU max
VRANGE : min THRU max
HPOSITION : no. of spaces (50)
VPOSITION : no. of lines (40)
TITLE : 'title'
VLABEL : 'vert-label'
HLABEL : 'horiz-label'
Hints:
1. Single quotes (') must appear around labels and titles.
2. PLOT ? puts STATMATE into subcommand mode.
3. PLOT alone puts STATMATE into subcommand mode.
STEPWISE
STEPWISE
Purpose: Performs a stepwise multiple regression on several variables
Examples:
STEPWISE Y ON X1,X2,TABLE=FIT,PAR,ANOVA,FENTER=4.1
STEP ?
Modifiers:
TABLE : (PARAM), (FIT), ANOVA, SEP or
ALL or NONE
MAXSTEP : (20)
FORCE : (0)
FENTER : (4.0)
FREMOVE : (3.8)
METHOD : (STEPWISE), BACKWARD, FORWARD
Notes:
1. STEP ? puts STATMATE into subcommand mode.
2. STEP alone puts STATMATE into subcommand mode.
CHART
CHART
Purpose: Calculates R- and X-control chart statistics
Examples:
CHART WIDTH ON MACHINE,TYPE=X,CONTROL=RANGE,TITLE='GAUGE DATA'
Modifiers:
KCENTER : (0.0) known value KSIGMA : (1.0) known value
BASIS : (SIGMA), RANGE, KNOWN TYPE : (XCHART), RCHART
DISPLAY : (TABLE,PLOT), NONE
HMARK : (5) HFILLER : (2)
VRANGE : (DATA) min THRU max VPOSITION : no. of lines (40)
TITLE : 'title' VLABEL : 'vert-label'
HLABEL : 'horiz-label'
Hints:
1. Use KCENTER and KSIGMA only when you want the
corresponding chart values to be interpreted from the
value specified, i.e., a (K)nown standard.
2. Note that the default on control limits is SIGMA. If
you use TYPE=RCHART, then the default of CONTROL=SIGMA is an
invalid request. In this case, you must use CONROL=RANGE
or CONTROL=KNOWN.
3. CHART ? avaialble
CUSUM
Purpose: Produces a cumulative sum (cusum) chart
Examples:
CUSUM WIDTH ON MACHINE,RESET=YES,HMARK=10,TARGET=0.45
Modifiers:
TARGET : (0.0)
DISPLAY : (TABLE,PLOT)
RESET : (NO) YES
HMARK : (5)
HFILLER : (2)
VRANGE : (DATA) or min THRU max
VPOSITION : no. of lines (40)
TITLE : 'title'
VLABEL : 'vert-label'
HLABEL : 'horiz-label'
BREAKDOWN
BREAKDOWN
Purpose: Produces simple statistics for groups of data
Examples:
BREAK HEIGHT ON SEX
BREAK PRODUCTION ON COMPANY
Hints:
1. Subgroups are defined by the data found in the second, group,
variable. Statistics are generated for each subgroup.
NONLINEAR
NONLINEAR
Purpose: Performs a nonlinear regression
Examples:
NONLINEAR MODEL=GOMPERTZ.FIL,CONV=.004,MAXITER=20,REPORT=5,TYPE=FIT
Modifiers:
MODEL : file name (MODEL)
CONVERGE : (.001)
MAXITER : (50)
REPORT : (5)
TYPE : (NONE), FIT, FORECAST, RESIDUAL
Hints:
1. MODEL file statements must be in the order: PARAM, INIT, MODEL.
For example,
param alpha,beta
initial 22.8, 37.44
model urbanpop=alpha+exp(beta*year)
2. Limited to at most 5 variables and 5 parameters each.
3. Terminate very slow running problems with Control-C key.
4. Use REPORT to see intermediate results.
5. The last line in the model file must be followed with a
carriage return.
6. When using TYPE, the SET OUT command must be used before
executing NONLINEAR.
RCORR
RCORRELATION
Purpose: Produces rank correlation coefficients
Examples:
RCORR VARA ON VARB,TEST=SPEARMAN,KENDALL
Modifiers:
TEST : (SPEARMAN) KENDALL
stop
Volume in drive A has no label
Directory of A:\
DEMO 5639 7-03-87 9:58a
FILES861 TXT 2246 2-08-89 2:00p
GO BAT 38 11-05-87 3:26p
GO TXT 540 2-08-89 12:06p
HALD DAT 512 5-03-87 6:18p
MOTOR DAT 2560 5-03-87 6:18p
QAMEAS DAT 512 5-03-87 6:19p
README 7475 1-28-89 2:01p
SAMPLE DAT 640 5-03-87 6:19p
SMASKC OVR 12288 9-15-87 9:17p
SMATE EXE 57344 9-15-87 9:33p
SMBRKD OVR 8192 9-15-87 9:23p
SMHELP TXT 16512 6-02-87 8:32p
SMINLX OVR 3072 9-15-87 9:18p
SMINPT OVR 9216 9-15-87 9:19p
SMINSTLL EXE 37376 2-27-80 11:37p
SMKOLM OVR 12800 9-15-87 9:28p
SMONET OVR 14336 9-15-87 9:26p
SMONEW OVR 6656 9-15-87 9:26p
SMONPR OVR 12288 9-15-87 9:28p
SMQDATA COM 19292 5-14-88 9:10p
SMRCOR OVR 15872 9-15-87 9:27p
SMSA$ 3456 9-15-87 9:09p
SMSETI OVR 5120 9-15-87 9:17p
SMSHOW OVR 5120 9-15-87 9:19p
SMSTAT OVR 9216 9-15-87 9:20p
SMSYAN OVR 13312 9-15-87 9:18p
SMTNPR OVR 11264 9-15-87 9:28p
SMTTES OVR 5120 9-15-87 9:21p
SMTWOT OVR 11776 9-15-87 9:27p
SMTWOW OVR 12800 9-15-87 9:27p
USPOP DAT 640 5-03-87 6:19p
USPOPDEM DAT 512 5-03-87 6:19p
33 file(s) 323742 bytes
27648 bytes free