USE a Data File in Blossom Statistics

The USE command specifies the data file to be used. The command line syntax of the USE command is:

            USE [data filename] [/ variable name list]

The simple USE statement (with no arguments following on the line) provides a list of files available (Console version) or a Windows file access dialog box. A filename may be specified with the form USE data filename. If the file is an ASCII text file and contains no variable names (labels), these should be added after the filename with /variable name list (a "slash" followed by variable names in the correct order and number as in the file).

Blossom determines whether the file being read by the USE command is an ASCII text, Data Interchange Format (DIF), S-Plus 2000 (*.), or SYSTAT (SYS or SYD) file. The USE command has two forms for ASCII text data files: USE filename and USE filename / variable list. The first form specifies a labeled and the second an unlabeled data file. (The structure of and differences between these files is described above in the Data Formats section.) For example,

        >USE STUDY.DAT

causes Blossom to read the labeled data file STUDY.DAT and provide a list of variables in the file and number of cases read. In this example, the period and file extension need not be entered since "DAT" is the default file extension for Blossom data files. Other file extensions must be supplied explicitly. Data files with no extension are indicated as such by entering the file name followed by a period (e.g., USE DATA.)

To use unlabeled data files specify the variable labels (names) after the name of the file. The command

        >USE FIELD1.DAT / GROUP PLOT RESPONS1 RESPONSE2

causes Blossom to read file FIELD1.DAT and assign labels (variable names) GROUP, PLOT, RESPONS1, and RESPONSE2 to the four data variables contained in the file. Labels can be entered in upper or lower case, but is always interpreted by Blossom as upper case. The number of variables in the list following the slash (/) of the USE command must match exactly the number of columns in the data file. Therefore, for example, to analyze only the first four variables in a data file containing six variables (columns), all six variable names must be entered. (Later, a subset of the variables can be specified within the statistical command line).

To read Data Interchange Format (DIF) and SYSTAT (.SYS and .SYD) data files, the entire filename including the extension is entered. Blossom assumes a data file is an ASCII text file with a DAT extension if no extension is provided. Variable names from SYSTAT and Data Interchange Format files are automatically read in by Blossom. The command:

        >USE GROUSE.SYS

reads in all variables from the SYSTAT file GROUSE.SYS.

Data in the USEd file are ready for statistical analysis and are available until another USE command is given.

The command:

        >USE

(without a filename specification) provides a list of all files in the local folder. The Console version prompts for input of the filename to USE. In the Windows version, this abbreviated command invokes the "Use Data File" dialog box.

In the Console version, a subset of all files is obtained by giving the USE command with a wildcard specification. For example the command:

        >USE *.SYS

provides a list of files with a .SYS extension, and the command:

        >USE BIRD*.*

lists files with any extension that begin with "BIRD".

In the Windows version a data file can be USEd by selecting "Use/Submit Files | Use Data File" and interacting with the "Use Data File" dialog box. The F2 Function Key or the "Use Dataset" button on the toolbar also invokes this dialog box. A drop-down selection list for "Files of Type" lists (all) Data files (files with extensions DAT, SYS, SYD, or DIF), SYSTAT datasets (files with extensions SYS or SYD), DIF Files (files with extension (DIF), or all files in the local folder.

In the Console version of Blossom, a data file can be USEd by giving the filename as an argument to the CONBLOS.EXE program name at the operating system command prompt. For example, the following Console version session invokes CONBLOS and USEs the BGROUSE.DAT data file:

        D:\Blossom\MyData\ElPaso>CONBLOS BGROUSE

    File being used is BGROUSE.DAT with 21 cases and 3 variables.
    The variables are: DIST, ELEV, SEX
          _
        _(_)_
       (_)@(_)
         (_)
      ._ |
       \)| _.
        \|(/
         |/
    .--- --- -.
    Blossom Version C2005.08.26
    Fort Collins Science Center
    U.S. Geological Survey
    2150 Centre AV BLDG C
    Fort Collins, CO 80526-8118, USA
    http://www.fort.usgs.gov/products/software/blossom/blossom.asp
    >

In addition, a statistical procedure to be performed can be specified with arguments to the CONBLOS invocation. For example, the following Console version session invokes CONBLOS, USEs the BGROUSE.DAT data file, and runs a multiresponse permutation procedure (MRPP) of variables DIST and ELEV using the grouping variable SEX:

        D:\Blossom\MyData\ElPaso>CONBLOS BGROUSE MRPP DIST ELEV * SEX

    File being used is BGROUSE.DAT with 21 cases and 3 variables.
    The variables are: DIST, ELEV, SEX
          _
        _(_)_
       (_)@(_)
         (_)
      ._ |
       \)| _.
        \|(/
         |/
    .--- --- -.
    Blossom Version C2005.08.26
    Fort Collins Science Center
    U.S. Geological Survey
    2150 Centre AV BLDG C
    Fort Collins, CO 80526-8118, USA
    http://www.fort.usgs.gov/products/software/blossom/blossom.asp


                     Multi-Response Permutation Procedure (MRPP)

     Data Used
                 Data File: BGROUSE.DAT
        Grouping Variable: SEX
       Response Variables: DIST, ELEV

    Specification of Analysis
       Number of observations: 21
             Number of groups: 2
            Distance exponent: 1.00000000000000
             Weighting factor: n(I)/sum(n(I)) = C(I) = 1

    Group Summary
      Group Value Group Size Group Distance
      3.00000000000000 9 1.07214652525827
      4.00000000000000 12 1.39643892970427

    Variable Commensuration Summary
       Variable Name Average Distance (Euclidean if V=1)
       DIST 9264.76190476191
       ELEV 279.228571428571

    Results
       Delta Observed = 1.25745647065599
       Delta Expected = 1.51256336315532
       Delta Variance = 0.270618755524092E-002
       Delta Skewness = -2.09758982732985

                   Standardized test statistic = -4.90391852737653
                      Probability (Pearson Type III) of a
                        smaller or equal delta = 0.298316800990588E-002

    Output was appended to file "BGROUSE.OUT"

    >