ProfitPress Mega CDROM2 …eeware (MSDOS)(1992)(Eng)/ProfitPress-MegaCDROM2.B6I/APPS/STATISTI/ESDISK.ZIP/READ.ME

Text File | 1990-01-24 | 16.5 KB | 415 lines

        EASISTAT DISK READ.ME FILE

        List of files on disk:

        1. READ.ME (this file).

        2. EASISTAT.EXE - the EASISTAT statistics program.

        3. EASISTAT.DOC - on-disk documentation explaining how to use
        EASISTAT.

        4. DEMO - an example input file which produces a demonstration
        of some of EASISTAT's features.

        5. EXAMPLE.DAT - an example data set used by DEMO.

        6. DEMO.BAT - a short batch file which allows you to run the
        demonstration just by typing DEMO.

        7. ORDINV.PRN - print off this file to obtain an order and
        invoice form.

        8. PRODUCTS - a list of products available from ARC Scientific.


        THE SHAREWARE AGREEMENT

        EASISTAT is distributed as a Shareware program. That means that
        you are legally allowed to copy it and give copies to other
        people who you think might like it - in fact you are
        encouraged to do so. However please note the following points:


        1. The copyright of the program remains with ARC Scientific
        Limited.

        2. You must not charge more than a nominal fee for the program
        if you pass it on to someone else, nor may you distribute the
        disc as part of a promotional scheme.

        3. Please acknowledge the program in any publications of work
        for which it was used.

        4. If you pass it on you must do so in unmodified form and the
        following eight files must all be passed on together:

         EASISTAT.EXE   EASISTAT.DOC   EXAMPLE.DAT  DEMO
         DEMO.BAT       READ.ME        ORDINV.PRN   PRODUCTS

        5. If you decide to use EASISTAT, please pay 30 pounds, the proper
        registration fee, to ARC Scientific Limited.  This gets you an
        up-to-date copy with a printed manual, and encourages the
        production of other good, low-cost software.


        For your convenience in registering, the file ORDINV.PRN can
        be used, with most printers, to print out an order form and/or
        and invoice; you may need to complete the latter in order to get
        your employer to pay the registration fee.



        UPDATES TO EASISTAT - v. 1.2

        The following alterations have been made to EASISTAT since the
        manual was printed and since the on-disk documentation was
        written:

        1. EASISTAT can now produce a variety of graphs using EASIPLOT,
        a Shareware graph-drawing program available from ARC Scientific
        Ltd. All references to PCG2 in the manual (the graph-drawing
        program supplied with PC-File) should be taken to refer to
        EASIPLOT instead. Graph files produced from the CHISQ function
        can still be displayed using PCG2 if desired, but several other
        functions can also be displayed graphically if EASIPLOT is used
        (all those appearing on the main menu followed by a 'G'). See
        the file EASIPLOT.DOC supplied with EASIPLOT for full details of
        the graph types available. EASIPLOT is supplied to all
        registered users of EASISTAT and is also available from Shareware
        libraries.

        2. Wilcoxon's signed rank test can be performed on a column by
        selecting the 'S' option (for 'signed'). To use the test to
        compare the values in two columns, use the DERIVE command to
        make a third column the difference between these two, then
        perform the test on the third column. E.g. to compare columns 1
        and 2:

        NEW C3
        DERIVE C3 C1-C2
        WILC S C3

        A one-tailed probability value for the two distributions to
        differ by such a magnitude and in such a direction is output.

        3. The Kruskal-Wallis one-way analysis of variance by ranks test
        is now available. It is selected instead of the normal
        ANOVA by using the 'N' option (for 'nonparametric'). The
        column and groups are then selected in the same way as for
        ANOVA, and the output reports the overall differences between
        the group ranks and between pairs of groups as described in
        Nonparametric Statistics for the Behavioural Sciences by Siegel.
        Example:

        ANOVA N C15
        5
        C5=1
        C5=2
        C5=3
        C5=4
        C5=5

        4. Confidence limits can now be displayed for the mean of a
        sample and the mean difference between two samples (for t
        tests), and for the gradient and correlation coefficient in the
        linear regression function. Because they slow the program down
        slightly they are not displayed by default. If you wish to have
        confidence limits calculated then when you first run EASISTAT
        enter:

        EASISTAT /L95

        The 'L' switch (for 'limits') should be followed immediately by
        any number between 5 and 99.9 which is the value of the
        confidence interval. For the duration of the session confidence
        limits will be calculated for the appropriate functions (i.e.
        BASICS, TTEST and REGRESS). This switch can be used with the any
        of the other switches, as described in the EASISTAT manual, and
        the name of a command file may also be given. Alternatively
        confidence limits can be selected during a session by using the
        new LIMITS command, e.g.:

        LIMITS 95

        To cease displaying confidence limits enter:

        LIMITS 0

        5. When data files are read in, columns are automatically
        formatted so that the correct number of decimal places is
        displayed for each column.

        6. Twenty general purpose variables are available which can be
        referred to in expressions in the same way as columns. They are
        named V1 to V20. The names can be changed using the LABEL
        command, e.g.:

        LABEL V3 PI

        Values can be assigned to the variables using the DERIVE
        command, e.g.:

        DERIVE V2 2.7
        DERIVE V3 ARCCOS(0-1)

        The values of the variables can be listed with the LIST
        VARIABLES command or just L V for short.

        7. The parser has been improved so that columns and variables
        can now be referred to by their titles instead of just c1, c2,
        etc. You can enter commands like:

        BASICS GHQ IF SEX=2

        REGRESS HDA HDD

        The bug in the parser involving operator precedence has now
        been fixed, so that 3-4+5 now equals 4, as it should. Operator
        precedence is as described in the EASISTAT manual, except that
        the following groups of operators have equal precedence (within
        each group) and are evaluated left to right:

        !=   =
        >    <    >=   <=
        -    +
        /    *
        PX   PT

        Additionally very large and very small numbers can be referred
        to using exponential notation, where the mantissa of the number
        is followed immediately by an 'E' and the power of ten to
        multiply it by. Thus 3.3e5 is the same as 33000 and 2.7e-4
        means 0.00027. The results of the ARITHMETIC function are
        output in this format.

        These improvements do mean that you must obey certain rules when
        naming columns and variables so that EASISTAT doesn't get confused:

        a) Don't start any title name with a digit e.g. 1stname, 2A,
        etc.

        b) Don't start any title name with a 'C' or a 'V' followed by a
        digit e.g. C2D, c40, v2, etc.

        c) Don't call any column 'ROW', 'P', 'G', 'E', 'RANKS' or 'ZED'.

        d) Don't start any title name with the name of a mathematical
        function e.g. sint (uses SIN), pname (uses pN), etc. See the
        full list of functions in the manual.

        e) Don't start any title name with the name of a previous
        title, i.e. if one column is called DA, then no column to the
        right of it must begin with the letters DA.

        f) Capital and lower case letters are treated as identical.

        8. When creating a new column, the name and optionally format
        can be given at the same time, e.g.:

        NEW C2 TIM
        NEW C1 FRED 6 4

        9. When you input a command file, it's possible to pass
        parameters to it, as with DOS batch files. If you have an
        input file called EXAMPLE that contains the following line:

        ECHO parm1 %1 parm2 %2 parm3 %3

        then entering:

        INPUT EXAMPLE one two testing

        will produce the following output:

        parm1 one parm2 two parm3 testing

        This is mainly to allow you to perform the same set of tests
        repeatedly on different columns.

        10. Using the BASICS command sets the value of three new
        variables called XTOTAL, XMEAN and XNUMBER (do not use these
        for column names either). These names can then be used in
        expressions to represent the total of all the values, the mean
        of all the values and the number of rows used in the last
        BASICS command, respectively. The main idea is to enable people
        to implement their own statistical tests. For example, if
        EASISTAT did not perform the Wilcoxon signed rank test as a
        built in function, one could do it "by hand" with a command
        file called WILCSR reading as follows:

        ECHO Command file "WILCSR" to perform signed rank test
        ECHO     comparing %1 and %2
        NEW C1 DIF
        DERIVE DIF %1 - %2
        NARROW %1 != %2
        DERIVE DIF ABS DIF
        DERIVE DIF RANK DIF
        BASICS DIF
        LABEL V1 N'
        LABEL V2 T1
        LABEL V3 T2
        LABEL V4 VAR
        DERIVE N' XNUMBER
        BASICS DIF IF %1>%2
        DERIVE T1 XTOTAL
        BASICS DIF IF %1<%2
        DERIVE T2 XTOTAL
        DERIVE VAR N'*(N'+1)*(2*N'+1)/24
        ECHO Standardised normal deviate is:
        ARITH (T1-0.25*N'*(N'+1)-0.5)/(VAR POW 0.5)
        ECHO
        ECHO P value (single-tailed) is:
        ARITH 1-PN((ABS(T1-0.25*N'*(N'+1))-0.5)/(VAR POW 0.5))
        DEL DIF

        Then to compare columns HDA and HDD one would just enter:

        INPUT WILCSR HDA HDD

        Don't be intimidated by the above, it's just copied straight
        out of a statistics textbook, and hopefully using the same
        principles you should be able to perform most simple tests that
        might be missing from EASISTAT.

        11. Entering command lines has been made easier. The LEFT and
        RIGHT cursor keys can be used to move along the line, and using
        the CONTROL key with them moves a word at a time. HOME and END
        move to the start or end of the line. Insert and overwrite mode
        are toggled with the INSERT key. BACKSPACE and DELETE keys
        delete a letter left and right. CONTROL HOME and CONTROL END
        delete to the start or finish of the line. The UP and DOWN
        cursor keys scroll you through previous commands to allow you to
        edit those. The ESCAPE key deletes the whole line.

        12. A new function called COPY to copy a block of data from one
        place to another has been added. To use it one enters coordinates
        of the top left corner and bottom right corner of the block to
        be copied from, and the coordinates of the top left corner of
        the destination block. These six numbers are entered on one
        line, separated by commas, e.g.:

        COPY 1,1,2,10,3,1

        copies the first two columns of the first ten rows into the
        third and fourth columns. This function may occasionally be
        useful to reorganise the layout of data. Note that no new rows
        or columns are created - the space for the destination block
        must already exist.

        13. A general purpose curve-fitting and function-minimisation
        routine has been added called MINIMISE. To use it you enter an
        arithmetic expression and then the names of the variables within
        that expression which can be altered to minimise the value of
        expression over all the data rows. Usually the aim will be to
        find the best fit of the expression to the values in one column
        and in this case the expression will automatically converted
        into the expression for the least-squares fit to that column. If
        the original function is to be minimised instead, then input
        NONE (or just N) instead of a column name.

        For example, to perform multiple linear regression with HDD as
        the dependent variable and A, B, C, and D as independent
        variables:

        MIN                                     << function name
        V1 + V2*A + V3*B + V4*C + V5*D          << expression to fit
        HDD                                     << column to fit to
        V1 V2 V3 V4 V5                          << variables which can
                                                   be adjusted

        The output will show you that the following function is in fact
        the one which is minimised:

        Sigma: ((V1 + V2*A + V3*B + V4*C + V5*D)-HDD)pow2

        This is the function for the least sum of squares difference
        between the supplied function and the column to fit to.

        General curve-fitting is slower, less accurate and less
        informative than the linear functions supplied so it is best to
        try to convert your function to a linear form instead. It is up
        to you to make sure that the function has a minimum, and to set
        appropriate starting values for the variables so that the global
        minimum is found if there is more than one local minimum.

        The iterative process stops when one step fails to reduce the
        absolute value of the function by one thousandth. This means
        that if the function has a value of 2 then the last step-size is
        less than 0.0002. Hoewever if the function (which may be the same
        shape) has a value of 200000 then the last step-size may be up to
        20. If you want higher accuracy then you will have to add a
        constant to the function which reduces its absolute value to
        close to zero (in the latter example one would add -200000) and
        start the minimisation process again.

        Here are some other examples:

        MIN 6*V1pow2+4*V1-123
        NONE
        V1

        MIN V1*C1*exp(C2powV2)
        C3
        V1 V2

        14. The three linear regression and curve-fitting routines allow
        you to specify an optional extra column into which will be put
        the predicted values from the best-fitting function found. This
        allows you to compare these values with the actual values of the
        dependent variable, or to subtract them if you wish to "partial
        out" the effects of the independent variables. For the REGRESS
        function the column for the predicted values is specified as a
        third column after the second and on the same line. For the
        MULTIPLE regression and MINIMISE a second column is specified
        after the column for the dependent variable. E.g.:

        REG HDA HDD PRED
                or:
        REG
        HDA
        HDD PRED

        MULT HDA PRED
        2
        A
        B
                or:
        MULT
        HDA PRED
        2
        A
        B

        MIN
        exp(V1*A)+exp(V2*B)
        HDD PRED
        V1 V2

        In each case the derived values will be entered into the column
        PRED.

        15. All the data rows and columns can be deleted by entering:

        DELETE ALL


                        SOME OF OUR OTHER PRODUCTS

        The file PRODUCTS gives you information about programs
        available from ARC Scientific Limited.  We also provide a
        consultancy service on scientific computing - ask us for more
        information if you would like some help.  Our other Shareware
        product, CPMDISK, can also be ordered using the forms in
        ORDINV.PRN.



           ARC Scientific Limited,
           257 Woodstock Road,
           Oxford OX2 7AE,
           ENGLAND.

           Telephone: [phone]