# NAG Library Routine Document

## 1Purpose

g02bhf computes means and standard deviations, sums of squares and cross-products of deviations from means, and Pearson product-moment correlation coefficients for selected variables omitting completely any cases with a missing observation for any variable (either over all variables in the dataset or over only those variables in the selected subset).

## 2Specification

Fortran Interface
 Subroutine g02bhf ( n, m, x, ldx, miss, kvar, xbar, std, ssp, r, ldr,
 Integer, Intent (In) :: n, m, ldx, mistyp, nvars, kvar(nvars), ldssp, ldr Integer, Intent (Inout) :: miss(m), ifail Integer, Intent (Out) :: ncases Real (Kind=nag_wp), Intent (In) :: x(ldx,m) Real (Kind=nag_wp), Intent (Inout) :: xmiss(m), ssp(ldssp,nvars), r(ldr,nvars) Real (Kind=nag_wp), Intent (Out) :: xbar(nvars), std(nvars)
#include nagmk26.h
 void g02bhf_ (const Integer *n, const Integer *m, const double x[], const Integer *ldx, Integer miss[], double xmiss[], const Integer *mistyp, const Integer *nvars, const Integer kvar[], double xbar[], double std[], double ssp[], const Integer *ldssp, double r[], const Integer *ldr, Integer *ncases, Integer *ifail)

## 3Description

The input data consists of $n$ observations for each of $m$ variables, given as an array
 $xij, i=1,2,…,n n≥2,j=1,2,…,m m≥2,$
where ${x}_{ij}$ is the $i$th observation on the $j$th variable, together with the subset of these variables, ${v}_{1},{v}_{2},\dots ,{v}_{p}$, for which information is required.
In addition, each of the $m$ variables may optionally have associated with it a value which is to be considered as representing a missing observation for that variable; the missing value for the $j$th variable is denoted by ${\mathit{xm}}_{j}$. Missing values need not be specified for all variables. The missing values can be utilized in two slightly different ways; you can indicate which scheme is required.
Firstly, let ${w}_{i}=0$ if observation $i$ contains a missing value for any of those variables in the set $1,2,\dots ,m$ for which missing values have been declared, i.e., if ${x}_{ij}={\mathit{xm}}_{j}$ for any $j$ ($j=1,2,\dots ,m$) for which an ${\mathit{xm}}_{j}$ has been assigned (see also Section 7); and ${w}_{i}=1$ otherwise, for $\mathit{i}=1,2,\dots ,n$.
Secondly, let ${w}_{i}=0$ if observation $i$ contains a missing value for any of those variables in the selected subset ${v}_{1},{v}_{2},\dots ,{v}_{p}$ for which missing values have been declared, i.e., if ${x}_{ij}={\mathit{xm}}_{j}$ for any $j$ ($j={v}_{1},{v}_{2},\dots ,{v}_{p}$) for which an ${\mathit{xm}}_{j}$ has been assigned (see also Section 7); and ${w}_{i}=1$ otherwise, for $\mathit{i}=1,2,\dots ,n$.
The quantities calculated are:
(a) Means:
 $x-j=∑i=1nwixij ∑i=1nwi , j=v1,v2,…,vp.$
(b) Standard deviations:
 $sj= ∑i= 1nwi xij-x-j 2 ∑i= 1nwi- 1 , j=v1,v2,…,vp.$
(c) Sums of squares and cross-products of deviations from means:
 $Sjk=∑i=1nwixij-x-jxik-x-k, j,k=v1,v2,…,vp.$
(d) Pearson product-moment correlation coefficients:
 $Rjk=SjkSjjSkk , j,k=v1,v2,…,vp.$
If ${S}_{jj}$ or ${S}_{kk}$ is zero, ${R}_{jk}$ is set to zero.

None.

## 5Arguments

1:     $\mathbf{n}$ – IntegerInput
On entry: $n$, the number of observations or cases.
Constraint: ${\mathbf{n}}\ge 2$.
2:     $\mathbf{m}$ – IntegerInput
On entry: $m$, the number of variables.
Constraint: ${\mathbf{m}}\ge 2$.
3:     $\mathbf{x}\left({\mathbf{ldx}},{\mathbf{m}}\right)$ – Real (Kind=nag_wp) arrayInput
On entry: ${\mathbf{x}}\left(\mathit{i},\mathit{j}\right)$ must be set to ${x}_{\mathit{i}\mathit{j}}$, the value of the $\mathit{i}$th observation on the $\mathit{j}$th variable, for $\mathit{i}=1,2,\dots ,n$ and $\mathit{j}=1,2,\dots ,m$.
4:     $\mathbf{ldx}$ – IntegerInput
On entry: the first dimension of the array x as declared in the (sub)program from which g02bhf is called.
Constraint: ${\mathbf{ldx}}\ge {\mathbf{n}}$.
5:     $\mathbf{miss}\left({\mathbf{m}}\right)$ – Integer arrayInput/Output
On entry: ${\mathbf{miss}}\left(j\right)$ must be set equal to $1$ if a missing value, $x{m}_{j}$, is to be specified for the $j$th variable in the array x, or set equal to $0$ otherwise. Values of miss must be given for all $m$ variables in the array x.
On exit: the array miss is overwritten by the routine, and the information it contained on entry is lost.
6:     $\mathbf{xmiss}\left({\mathbf{m}}\right)$ – Real (Kind=nag_wp) arrayInput/Output
On entry: ${\mathbf{xmiss}}\left(j\right)$ must be set to the missing value, $x{m}_{j}$, to be associated with the $j$th variable in the array x, for those variables for which missing values are specified by means of the array miss (see Section 7).
On exit: the array xmiss is overwritten by the routine, and the information it contained on entry is lost.
7:     $\mathbf{mistyp}$ – IntegerInput
On entry: indicates the manner in which missing observations are to be treated.
${\mathbf{mistyp}}=1$
A case is excluded if it contains a missing value for any of the variables $1,2,\dots ,m$.
${\mathbf{mistyp}}=0$
A case is excluded if it contains a missing value for any of the $p\left(\le m\right)$ variables specified in the array kvar.
8:     $\mathbf{nvars}$ – IntegerInput
On entry: $p$, the number of variables for which information is required.
Constraint: $2\le {\mathbf{nvars}}\le {\mathbf{m}}$.
9:     $\mathbf{kvar}\left({\mathbf{nvars}}\right)$ – Integer arrayInput
On entry: ${\mathbf{kvar}}\left(\mathit{j}\right)$ must be set to the column number in x of the $\mathit{j}$th variable for which information is required, for $\mathit{j}=1,2,\dots ,p$.
Constraint: $1\le {\mathbf{kvar}}\left(\mathit{j}\right)\le {\mathbf{m}}$, for $\mathit{j}=1,2,\dots ,p$.
10:   $\mathbf{xbar}\left({\mathbf{nvars}}\right)$ – Real (Kind=nag_wp) arrayOutput
On exit: the mean value, of ${\stackrel{-}{x}}_{\mathit{j}}$, of the variable specified in ${\mathbf{kvar}}\left(\mathit{j}\right)$, for $\mathit{j}=1,2,\dots ,p$.
11:   $\mathbf{std}\left({\mathbf{nvars}}\right)$ – Real (Kind=nag_wp) arrayOutput
On exit: the standard deviation, ${s}_{\mathit{j}}$, of the variable specified in ${\mathbf{kvar}}\left(\mathit{j}\right)$, for $\mathit{j}=1,2,\dots ,p$.
12:   $\mathbf{ssp}\left({\mathbf{ldssp}},{\mathbf{nvars}}\right)$ – Real (Kind=nag_wp) arrayOutput
On exit: ${\mathbf{ssp}}\left(\mathit{j},\mathit{k}\right)$ is the cross-product of deviations, ${S}_{\mathit{j}\mathit{k}}$, for the variables specified in ${\mathbf{kvar}}\left(\mathit{j}\right)$ and ${\mathbf{kvar}}\left(\mathit{k}\right)$, for $\mathit{j}=1,2,\dots ,p$ and $\mathit{k}=1,2,\dots ,p$.
13:   $\mathbf{ldssp}$ – IntegerInput
On entry: the first dimension of the array ssp as declared in the (sub)program from which g02bhf is called.
Constraint: ${\mathbf{ldssp}}\ge {\mathbf{nvars}}$.
14:   $\mathbf{r}\left({\mathbf{ldr}},{\mathbf{nvars}}\right)$ – Real (Kind=nag_wp) arrayOutput
On exit: ${\mathbf{r}}\left(\mathit{j},\mathit{k}\right)$ is the product-moment correlation coefficient, ${R}_{\mathit{j}\mathit{k}}$, between the variables specified in ${\mathbf{kvar}}\left(\mathit{j}\right)$ and ${\mathbf{kvar}}\left(\mathit{k}\right)$, for $\mathit{j}=1,2,\dots ,p$ and $\mathit{k}=1,2,\dots ,p$.
15:   $\mathbf{ldr}$ – IntegerInput
On entry: the first dimension of the array r as declared in the (sub)program from which g02bhf is called.
Constraint: ${\mathbf{ldr}}\ge {\mathbf{nvars}}$.
16:   $\mathbf{ncases}$ – IntegerOutput
On exit: the number of cases actually used in the calculations (when cases involving missing values have been eliminated).
17:   $\mathbf{ifail}$ – IntegerInput/Output
On entry: ifail must be set to $0$, $-1\text{​ or ​}1$. If you are unfamiliar with this argument you should refer to Section 3.4 in How to Use the NAG Library and its Documentation for details.
For environments where it might be inappropriate to halt program execution when an error is detected, the value $-1\text{​ or ​}1$ is recommended. If the output of error messages is undesirable, then the value $1$ is recommended. Otherwise, if you are not familiar with this argument, the recommended value is $0$. When the value $-\mathbf{1}\text{​ or ​}\mathbf{1}$ is used it is essential to test the value of ifail on exit.
On exit: ${\mathbf{ifail}}={\mathbf{0}}$ unless the routine detects an error or a warning has been flagged (see Section 6).

## 6Error Indicators and Warnings

If on entry ${\mathbf{ifail}}=0$ or $-1$, explanatory error messages are output on the current error message unit (as defined by x04aaf).
Errors or warnings detected by the routine:
${\mathbf{ifail}}=1$
On entry, ${\mathbf{n}}=〈\mathit{\text{value}}〉$.
Constraint: ${\mathbf{n}}\ge 2$.
${\mathbf{ifail}}=2$
On entry, ${\mathbf{nvars}}=〈\mathit{\text{value}}〉$ and ${\mathbf{m}}=〈\mathit{\text{value}}〉$.
Constraint: ${\mathbf{nvars}}\ge 2$ and ${\mathbf{nvars}}\le {\mathbf{m}}$.
${\mathbf{ifail}}=3$
On entry, ${\mathbf{ldr}}=〈\mathit{\text{value}}〉$ and ${\mathbf{nvars}}=〈\mathit{\text{value}}〉$.
Constraint: ${\mathbf{ldr}}\ge {\mathbf{nvars}}$.
On entry, ${\mathbf{ldssp}}=〈\mathit{\text{value}}〉$ and ${\mathbf{nvars}}=〈\mathit{\text{value}}〉$.
Constraint: ${\mathbf{ldssp}}\ge {\mathbf{nvars}}$.
On entry, ${\mathbf{ldx}}=〈\mathit{\text{value}}〉$ and ${\mathbf{n}}=〈\mathit{\text{value}}〉$.
Constraint: ${\mathbf{ldx}}\ge {\mathbf{n}}$.
${\mathbf{ifail}}=4$
On entry, $\mathit{i}=〈\mathit{\text{value}}〉$, ${\mathbf{kvar}}\left(\mathit{i}\right)=〈\mathit{\text{value}}〉$ and ${\mathbf{m}}=〈\mathit{\text{value}}〉$.
Constraint: $1\le {\mathbf{kvar}}\left(\mathit{i}\right)\le {\mathbf{m}}$.
${\mathbf{ifail}}=5$
On entry, ${\mathbf{mistyp}}=〈\mathit{\text{value}}〉$.
Constraint: ${\mathbf{mistyp}}=0$ or $1$.
${\mathbf{ifail}}=6$
After observations with missing values were omitted, no cases remained.
${\mathbf{ifail}}=7$
After observations with missing values were omitted, only one case remained.
${\mathbf{ifail}}=-99$
See Section 3.9 in How to Use the NAG Library and its Documentation for further information.
${\mathbf{ifail}}=-399$
Your licence key may have expired or may not have been installed correctly.
See Section 3.8 in How to Use the NAG Library and its Documentation for further information.
${\mathbf{ifail}}=-999$
Dynamic memory allocation failed.
See Section 3.7 in How to Use the NAG Library and its Documentation for further information.

## 7Accuracy

g02bhf does not use additional precision arithmetic for the accumulation of scalar products, so there may be a loss of significant figures for large $n$.
You are warned of the need to exercise extreme care in your selection of missing values. g02bhf treats all values in the inclusive range $\left(1±{0.1}^{\left({\mathbf{x02bef}}-2\right)}\right)×{xm}_{j}$, where ${\mathit{xm}}_{j}$ is the missing value for variable $j$ specified in xmiss.
You must therefore ensure that the missing value chosen for each variable is sufficiently different from all valid values for that variable so that none of the valid values fall within the range indicated above.

## 8Parallelism and Performance

g02bhf is not threaded in any implementation.

The time taken by g02bhf depends on $n$ and $p$, and the occurrence of missing values.
The routine uses a two-pass algorithm.

## 10Example

This example reads in a set of data consisting of five observations on each of four variables. Missing values of $0.0$ are declared for the second and fourth variables; no missing values are specified for the first and third variables. The means, standard deviations, sums of squares and cross-products of deviations from means, and Pearson product-moment correlation coefficients for the fourth, first and second variables are then calculated and printed, omitting completely all cases containing missing values for these three selected variables; cases $3$ and $4$ are therefore eliminated, leaving only three cases in the calculations.

### 10.1Program Text

Program Text (g02bhfe.f90)

### 10.2Program Data

Program Data (g02bhfe.d)

### 10.3Program Results

Program Results (g02bhfe.r)

© The Numerical Algorithms Group Ltd, Oxford, UK. 2017