The RAREEVENTS Procedure(Experimental)

Input Data Sets

Subsections:

The RAREEVENTS procedure accepts a single primary input data set of either of two types:

  • A DATA= data set contains process measurements to be analyzed.

  • A TABLE= data set contains a summary of a rare events chart, which consists of the measurements, probability limits, and other information.

These options are mutually exclusive. If you do not specify an option that identifies a primary input data set, PROC RAREEVENTS uses the most recently created SAS data set as a DATA= data set. Valid process measurements are greater than or equal to zero. Missing and negative values are ignored.

You can also specify a LIMITS= data set that contains probability limits for a rare events chart.

DATA= Data Set

A DATA= data set must include a process variable that contains measurements of the times between rare events. These measurements can be integers (for example, a count of days between events) or continuous values. In addition to the process variable, a DATA= data set can include the following:

  • BY variables

  • ID variables

  • index variable

The values of the optional index variable are used to label the horizontal axis tick marks on a rare events chart that is produced by a CHART statement. The index variable has no application in a COMPARE statement.

LIMITS= Data Set

A LIMITS= data set contains probability limit information for a rare events chart. Usually, you create a LIMITS= data set by specifying the OUTLIMITS= option in a CHART statement. You can use a LIMITS= data set to specify historical probability limits for a process or custom probability limits that are computed by other means.

Table 16.3 lists the variables that a LIMITS= data set contains.

Table 16.3: LIMITS= Data Set Variables

Variable

Description

_ALPHALPL_

Probability associated with the lower probability limit

_ALPHAUPL_

Probability associated with the upper probability limit

_DIST_

Name of the distribution used to compute the probability limits

_INDEX_

Name of the optional index variable

_LPL_

Lower probability limit

_MEDIAN_

Median of the probability distribution

_P_

Probability of success in a single Bernoulli trial on which the geometric distribution is based

_SHIFT_

Minimum possible value for the geometric distribution

_SIGMA_

Scale parameter for the exponential distribution

_THETA_

Threshold parameter for the exponential distribution

_TYPE_

Specifies whether the probability limits are estimated or standard

_UPL_

Upper probability limit

_VAR_

Name of the process variable that contains measurements of times between events


TABLE= Data Set

A TABLE= data set contains a summary of a rare events chart. Usually, you create a TABLE= data set by specifying the OUTTABLE= option in a CHART statement. You can use a TABLE= data set to display a previously created rare events chart or to specify custom probability limits by computing your own _LPL_ and _UPL_ values.

Table 16.4 lists the variables that a TABLE= data set contains.

Table 16.4: TABLE= Data Set Variables

Variable

Description

_ALPHALPL_

Probability associated with the lower probability limit

_ALPHAUPL_

Probability associated with the upper probability limit

_DIST_

Name of the distribution used to compute the probability limits

_EXLIM_

Flag that indicates that a probability limit was exceeded

index

Optional index variable

_LPL_

Lower probability limit

_MEDIAN_

Median of the probability distribution

process

Process variable containing measurements of times between events

_UPL_

Upper probability limit