Types of Reports Available

Four classes of reports are available in MICF. Samples of each report are shown in the MICF Reference.
micsrm140
Graphics reports are suitable for presentations and for helping with the recognition of trends and anomalies:
  • Color graphics require the use of a color terminal, pen plotter, or color printer.
  • Printer graphics produce black and white graphs using standard printers.
Printed reports list the data that you select in tables that can be used for reference purposes or further analyses. Statistical analyses offer a wide range of analytical tools that display your data and summarize its statistical characteristics.

Color Graphics

MICF offers a broad set of color graphics reports, including pie charts, block charts, star charts, simple X compared to Y plots, and complex plots of multiple variables. The SAS/GRAPH software product is required for generating these reports. The following color graphics options are available:
Vertical Bar Charts
MICF uses the VBAR option of SAS PROC GCHART to create color vertical bar (or column) charts.
Horizontal Bar Charts
MICF uses the HBAR option of SAS PROC GCHART to create color horizontal bar (or histogram) charts.
Pie Charts
MICF uses the PIE option of SAS PROC GCHART to create color pie (or circle) charts. By using optional parameters, you can generate groups of related pie charts on one page.
Block Charts
MICF uses the BLOCK option of SAS PROC GCHART to create color block (or three-dimensional column) charts.
Star Charts
MICF uses the STAR option of SAS PROC GCHART to create color star (or Kiviat) charts.
Color Plots
MICF uses SAS PROC GPLOT to create plots of one variable against another (for example, plot of response times by the number of page-ins per CICS transaction). By using optional parameters, you can create plots of multiple groups on a single axis (for example, plot of average number of users by hour for System 1 and plot of average number of users by hour for System 2).
Relationship Plots
MICF uses SAS PROC GPLOT to create scatter plots of two variables with a regression line. Each point on the plot represents the intersection of the values of the two variables and the regression line describes the relationship between the two variables.
Stacked Plots
MICF uses the SAS PROC GPLOT y*x=z option to create multiple plots on a single axis and scale with a detailed plot legend. For example, plot demand pages per second and system pages per second by hour of day. You have the option of stacking the plot lines (that is, plotting demand pages per second as the first line and the sum of demand pages per second and system pages per second -- total pages per second -- as the second line) or plotting the values exactly as they are on the database (that is, show the relation between demand pages per second and system pages per second rather than the sum).
Overlay Plots
MICF uses the SAS PROC GPLOT OVERLAY option to create multiple plots on a single axis and scale. For example, plot demand pages per second and system pages per second by hour of day. Overlay plots do NOT have a legend.

Printer Graphics

You can use MICF printer graphics to create an extensive variety of black and white hardcopy reports. These reports are useful in initial data analysis and if you do not have access to color graphics devices. The following printer graphics options are available:
Vertical Bar Charts
MICF uses the VBAR option of SAS PROC CHART to create vertical bar (or column) charts.
Horizontal Bar Charts
MICF uses the HBAR option of SAS PROC CHART to create horizontal bar (or histogram) charts.
Pie Charts
MICF uses the PIE option of SAS PROC CHART to create pie (or circle) charts.
Block Charts
MICF uses the BLOCK option of SAS PROC CHART to create block (or 3-dimensional column) charts.
Star Charts
MICF uses the STAR option of SAS PROC CHART to create star (or Kiviat) charts.
Printer Plots
MICF uses SAS PROC PLOT to create plots of one variable against another (for example, plot of response times by the number of page-ins per CICS transaction). By using optional parameters, you can create plots of multiple groups on a single axis (for example, plot of average number of users by hour for System 1 and plot of average number of users by hour for System 2).
Overlay Plots
MICF uses the SAS PROC PLOT OVERLAY option to create multiple plots on a single axis and scale (for example, plot demand pages per second and system pages per second by hour of day). As with printer plots, you can also create plots of multiple groups of data on the same axis (for example, plot of demand pages per second and system pages per second for System 1 and a plot of the same information for System 2).

Printed Reports

MICF printed reports let you list the data elements that you select in various convenient formats. The following printed reports options are available:
Standard List
MICF uses SAS PROC PRINT to create a simple listing of your data with title, subtitle, page number, inquiry name, and run date at the top of each page. Column headings are repeated on each page for ease of reference.
Structured List
MICF uses SAS PROC PRINT to create a listing of your data similar to a standard list but with indented format for subheadings in the body of your report. You can
Ranking Report
MICF uses SAS PROC PRINT to create a report similar to the standard list except that the observations are ordered by the value of a specific element. You can specify that either the top or bottom n observations be printed on the report. For example, as part of a job tuning effort, you could list the top 10 jobs based on cost.
Tabular Report
MICF creates a formal printed report with columns of data values (for example, response time, cost) over rows of data observations or time periods (for example, hours of the day). You specify report sequence, data elements to report, page break, subtotal controls, and report subheadings. You can request both subtotals and grand totals, with subtotal options including sum (accumulate), maximum/minimum value, percent of total, and percent/amount of change from the first observation. A subtotal derivation extended option is available to define your own derived subtotal values (for example, average response time). Optional parameters allow you to generate multiple report lines for each input file observation and to control the specific column in which each item is printed. Column headings are specified for each data value or overridden with free-form, multiple line column headings. You can even include footnotes on the report.
Tabular Ranking Report
MICF creates a formal report listing the top or bottom n observations of the input file that is based on the specified ranking element. The ranking is presented in Tabular Report format with columns of data values (for example, response time, cost) over rows of ranked observations (for example, userid). The listing includes selected data elements from each file observation and is useful in identifying areas for further study. You can optionally input detail observations, base the ranking on a user-specified summarization of these detail observations, and print the detail observations comprising the ranked summary observations. Summarization options include sum (accumulate), maximum/minimum value, percent of total, and percent/amount of change from the first observation. A subtotal derivation extended option is available to define your own derived subtotal values (for example, average response time). Optional parameters allow you to generate multiple report lines for each input file observation and to control the specific column in which each item is printed. Column headings are specified for each data value or overridden with free-form, multiple line column headings. You can even include footnotes on the report.

Statistical Analyses

MICF allows you to do most common descriptive and inferential statistical analyses on data from the
MICS database
. These analyses are useful in numerous performance, capacity planning, and installation accounting studies. The statistical analysis options are:
Frequency Distribution
The frequency distribution option generates a simple frequency distribution for a single variable. This option also reports frequency count, percentage, cumulative frequency, and cumulative percentage for each value of the variable. This option uses SAS PROC FREQ.
Crosstabulation Report
The crosstabulation report option generates a table that shows the combined frequency distribution for two variables. For example, a table for the variables SYSID and COMMAND would create a table with a separate column for each SYSID and a separate row for each TSO command that is contained in the data. The cells of the table would show the frequency with which each command was used on each system. In addition, cells can optionally contain the percentage of the observation that is represented by this cell, the percentage of the column total represented by the cell, and the percentage of the row total that is represented by the cell. This option uses SAS PROC FREQ.
3-Way Crosstabulation Report
The 3-way crosstabulation report option defines a set of tables that show the combined frequency distribution of three variables. For each value of Z, a 2-way crosstabulation report is created showing the combined frequency distribution for X and Y. For example, you might create a variable USERGRP that groups your systems users into a few categories. You might then create a 3-way crosstabulation showing SYSID by COMMAND for each value of USERGRP. Such an analysis could be useful in determining who is in need of training or who is making the greatest use of specific resources. This option uses SAS PROC FREQ.
N-Way Crosstabulation Report
The N-way crosstabulation report is a generalization of the 3-way crosstabulation report that allows you to show the combined frequency distributions for up to six variables. This option uses SAS PROC FREQ.
Descriptive Statistics Report
Use the descriptive statistics report option to generate basic descriptive statistics for numeric variables. The report can include:
  • Name of the variable
  • Number of observations
  • Sum of the weights of the observations
  • Sum of the observations
  • Mean
  • Standard deviation
  • Variance
  • Skewness
  • Kurtosis
  • Uncorrected sum of squares
  • Corrected sum of squares
  • Coefficient of variation
  • Standard error of the mean
  • Student's t value
  • Probability of a greater absolute value for Student's t
  • Approximation of the probability of a greater absolute value for student's t if the population mean is 0
  • Number of nonzero observations
  • Largest value
  • Smallest value
  • Range
  • Difference between the upper and lower quartile
  • Mode
  • Quartiles
  • 1st, 5th, 10th, 90th, 95th, and 99th percentile
  • Five lowest and highest observations
In addition, several statistics on missing values are provided. You can also request that several analytic plots be printed. This option uses SAS PROC UNIVARIATE.
Correlation Report
The correlation report option lets you examine the linear relationship between a pair of variables by generating several correlation statistics. The analysis can be extended to up to five variables that are correlated against five different variables. Statistics that you can request include:
  • Pearson correlations
  • Spearman coefficients
  • Kendall tau-b coefficient
  • Hoeffding's D statistic
  • Descriptive statistics
  • Significance probabilities
  • Sums of squares cross product matrices
  • Covariance matrix
You can specify one of four variance calculations and whether to accept missing values. This option uses SAS PROC CORR.
Stepwise Regression Report
The stepwise regression report option produces a model for a dependent variable Y based on several independent variables X1, X2, ... , Xn. You can select either forward or backward regression and can specify the minimum and maximum R-square values for inclusion (or exclusion). An analysis of variance report is produced for each step in the procedure. You can also define entry significance levels and stay levels to be used when building your model. You can force selected entries to be used in the model and to specify the minimum and maximum number of variables to be used in the model. Model building statistics are shown if you request them. This option uses SAS PROC STEPWISE.