Data functions

 

The data operators described in this section all work independently of the current data selection. Thus, each data operator takes into account the maximum possible number of data cases.

Individual data cases can be accessed in equations by giving the name of the data carrier together with the case number in parenthesis - this is the same format as when the data is being defined. For example, we could have

BD>data : Patientage(3)=18

BD>print : (Patientage(3))

The argument supplied in parenthesis for the observation number may be any valid equation, which will be rounded to what should be a positive integer. A simple example is as follows.

BD>print : (Patientage(1+2))

It will be an error if the ``observation number'' thus supplied does not correspond to an observation. The other functions described in this section generally deal with all of the data defined over particular data carriers, rather than individual cases.

Data percentiles

 

Usage

• centile  operator; two operands, D followed by E, in parenthesis and separated by a comma. D is a data carrier. E is any valid equation. The result of the equation, when rounded, should be an integer I in (0,100).

The centile  operator gives the Ith percentile of currently selected data values on D. For example, if we order the data values as , then is the smallest value such that at least of the values are less than or equal to this value, and at least of the values are greater than or equal to this value. The lower of two candidates is returned. Note that placing gives the minimum, and gives the maximum. All possible data cases defined for the data carrier D are taken into account.

An example of the use of the command is as follows. Suppose that Temperature is the name of a data carrier, and that we require the median. (We generate the 50 artificially to illustrate the use of equations here).

BD>print : (centile (temperature,25+25))

 

Sample correlations

 

Usage

• scorr  operator; two arguments and in parenthesis and separated by a comma. and are data carriers.

This returns the sample correlation between and . All possible data cases defined both for the data carrier and the data carrier are taken into account.  

Sample covariances

 

Usage

• scov  operator; two arguments and in parenthesis and separated by a comma. and are data carriers.

This returns the sample covariance between and as times the sum of cross product deviations. All possible data cases defined both for the data carrier and the data carrier are taken into account.  

Highest data case

 

Usage

• maxcase  operator; one operand D in parenthesis. D is a data carrier.

This returns the largest case number for which there is data on the data carrier D. All possible data cases defined for the data carrier D are taken into account.  

Number of matching data cases

 

Usage

• match  operator; two arguments and in parenthesis and separated by a comma. and are data carriers.

This returns the number of cases for which there are observations on both and . All possible data cases defined both for the data carrier and the data carrier are taken into account. The match  operator is the bivariate analogue of the number  operator.  

Sample mean

 

Usage

• mean  operator; one operand D in parenthesis. D is a data carrier.

This returns the arithmetic average of the observations defined for D. All possible data cases defined for the data carrier D are taken into account. If the total number of cases is zero, or if D has not been defined, then the operator returns a value of zero and an error is reported.  

Number of data cases

 

Usage

• number  operator; one operand D in parenthesis. D is a data carrier.

returns the total number of observations for the data carrier . All possible data cases defined for the data carrier D are taken into account. The number  operator is the univariate analogue of the match  operator.  

Sample variances

 

Usage

• svar  operator; one operand D in parenthesis. D is a data carrier.

This returns the sample variance of D, as times the sum of squared deviations. All possible data cases defined for the data carrier D are taken into account.