Integer type:  int32  int64  nag_int  show int32  show int32  show int64  show int64  show nag_int  show nag_int

Chapter Contents
Chapter Introduction
NAG Toolbox

# NAG Toolbox: nag_correg_coeffs_zero_subset_miss_pair (g02bm)

## Purpose

nag_correg_coeffs_zero_subset_miss_pair (g02bm) computes means and standard deviations, sums of squares and cross-products about zero, and correlation-like coefficients for selected variables omitting cases with missing values from only those calculations involving the variables for which the values are missing.

## Syntax

[xbar, std, sspz, rz, ncases, cnt, ifail] = g02bm(x, miss, xmiss, kvar, 'n', n, 'm', m, 'nvars', nvars)
[xbar, std, sspz, rz, ncases, cnt, ifail] = nag_correg_coeffs_zero_subset_miss_pair(x, miss, xmiss, kvar, 'n', n, 'm', m, 'nvars', nvars)
Note: the interface to this routine has changed since earlier releases of the toolbox:
Mark 22: n has been made optional
.

## Description

The input data consists of n$n$ observations for each of m$m$ variables, given as an array
 [xij],  i = 1,2, … ,n (n ≥ 2),j = 1,2, … ,m (m ≥ 2), $[xij], i=1,2,…,n (n≥2),j=1,2,…,m (m≥2),$
where xij${x}_{ij}$ is the i$i$th observation on the j$j$th variable, together with the subset of these variables, v1,v2,,vp${v}_{1},{v}_{2},\dots ,{v}_{p}$, for which information is required.
In addition, each of the m$m$ 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$j$th variable is denoted by xmj${\mathit{xm}}_{j}$. Missing values need not be specified for all variables.
Let wij = 0${w}_{\mathit{i}\mathit{j}}=0$, if the i$\mathit{i}$th observation for the j$\mathit{j}$th variable is a missing value, i.e., if a missing value, xmj${\mathit{xm}}_{\mathit{j}}$, has been declared for the j$\mathit{j}$th variable, and xij = xmj${x}_{\mathit{i}\mathit{j}}={\mathit{xm}}_{\mathit{j}}$ (see also Section [Accuracy]); and wij = 1${w}_{\mathit{i}\mathit{j}}=1$ otherwise, for i = 1,2,,n$\mathit{i}=1,2,\dots ,n$ and j = 1,2,,m$\mathit{j}=1,2,\dots ,m$.
The quantities calculated are:
(a) Means:
 xj = ( ∑ i = 1nwijxij)/( ∑ i = 1nwij),  j = v1,v2, … ,vp. $x-j=∑i=1nwijxij ∑i=1nwij , j=v1,v2,…,vp.$
(b) Standard deviations:
 sj = sqrt(( ∑ i = 1nwij(xij − xj)2)/( ∑ i = 1nwij − 1)),   j = v1,v2, … ,vp. $sj=∑i= 1nwij (xij-x-j) 2 ∑i= 1nwij- 1 , j=v1,v2,…,vp.$
(c) Sums of squares and cross-products about zero:
 n S̃jk = ∑ wijwikxijxik,  j,k = v1,v2, … ,vp. i = 1
$S~jk=∑i=1nwijwikxijxik, j,k=v1,v2,…,vp.$
(d) Correlation-like coefficients:
 R̃jk = (S̃jk)/(sqrt(S̃jj(k) S̃kk(j))),   j,k = v1,v2, … ,vp , $R~jk=S~jkS~jj(k) S~kk(j) , j,k=v1,v2,…,vp ,$
where jj(k) = i = 1nwijwikxij2${\stackrel{~}{S}}_{jj\left(k\right)}=\sum _{i=1}^{n}{w}_{ij}{w}_{ik}{x}_{ij}^{2}$ and kk(j) = i = 1nwikwijxik2${\stackrel{~}{S}}_{kk\left(j\right)}=\sum _{i=1}^{n}{w}_{ik}{w}_{ij}{x}_{ik}^{2}$
(i.e., the sums of squares about zero are based on the same set of observations as are used in the calculation of the numerator).
If jj(k)${\stackrel{~}{S}}_{jj\left(k\right)}$ or kk(j)${\stackrel{~}{S}}_{kk\left(j\right)}$ is zero, jk${\stackrel{~}{R}}_{jk}$ is set to zero.
(e) The number of cases used in the calculation of each of the correlation-like coefficients:
 n cjk = ∑ wijwik,  j,k = v1,v2, … ,vp. i = 1
$cjk=∑i=1nwijwik, j,k=v1,v2,…,vp.$
(The diagonal terms, cjj${c}_{\mathit{j}\mathit{j}}$, for j = 1,2,,n$\mathit{j}=1,2,\dots ,n$, also give the number of cases used in the calculation of the means xj${\stackrel{-}{x}}_{\mathit{j}}$ and the standard deviations sj${s}_{\mathit{j}}$.)

None.

## Parameters

### Compulsory Input Parameters

1:     x(ldx,m) – double array
ldx, the first dimension of the array, must satisfy the constraint ldxn$\mathit{ldx}\ge {\mathbf{n}}$.
x(i,j)${\mathbf{x}}\left(\mathit{i},\mathit{j}\right)$ must be set to xij${x}_{\mathit{i}\mathit{j}}$, the value of the i$\mathit{i}$th observation on the j$\mathit{j}$th variable, for i = 1,2,,n$\mathit{i}=1,2,\dots ,n$ and j = 1,2,,m$\mathit{j}=1,2,\dots ,m$.
2:     miss(m) – int64int32nag_int array
m, the dimension of the array, must satisfy the constraint m2${\mathbf{m}}\ge 2$.
miss(j)${\mathbf{miss}}\left(j\right)$ must be set equal to 1$1$ if a missing value, xmj$x{m}_{j}$, is to be specified for the j$j$th variable in the array x, or set equal to 0$0$ otherwise. Values of miss must be given for all m$m$ variables in the array x.
3:     xmiss(m) – double array
m, the dimension of the array, must satisfy the constraint m2${\mathbf{m}}\ge 2$.
xmiss(j)${\mathbf{xmiss}}\left(j\right)$ must be set to the missing value, xmj$x{m}_{j}$, to be associated with the j$j$th variable in the array x, for those variables for which missing values are specified by means of the array miss (see Section [Accuracy]).
4:     kvar(nvars) – int64int32nag_int array
nvars, the dimension of the array, must satisfy the constraint 2nvarsm$2\le {\mathbf{nvars}}\le {\mathbf{m}}$.
kvar(j)${\mathbf{kvar}}\left(\mathit{j}\right)$ must be set to the column number in x of the j$\mathit{j}$th variable for which information is required, for j = 1,2,,p$\mathit{j}=1,2,\dots ,p$.
Constraint: 1kvar(j)m$1\le {\mathbf{kvar}}\left(\mathit{j}\right)\le {\mathbf{m}}$, for j = 1,2,,p$\mathit{j}=1,2,\dots ,p$.

### Optional Input Parameters

1:     n – int64int32nag_int scalar
Default: The first dimension of the array x.
n$n$, the number of observations or cases.
Constraint: n2${\mathbf{n}}\ge 2$.
2:     m – int64int32nag_int scalar
Default: The dimension of the arrays miss, xmiss and the second dimension of the array x. (An error is raised if these dimensions are not equal.)
m$m$, the number of variables.
Constraint: m2${\mathbf{m}}\ge 2$.
3:     nvars – int64int32nag_int scalar
Default: The dimension of the array kvar.
p$p$, the number of variables for which information is required.
Constraint: 2nvarsm$2\le {\mathbf{nvars}}\le {\mathbf{m}}$.

### Input Parameters Omitted from the MATLAB Interface

ldx ldsspz ldrz ldcnt

### Output Parameters

1:     xbar(nvars) – double array
The mean value, xj${\stackrel{-}{x}}_{\mathit{j}}$, of the variable specified in kvar(j)${\mathbf{kvar}}\left(\mathit{j}\right)$, for j = 1,2,,p$\mathit{j}=1,2,\dots ,p$.
2:     std(nvars) – double array
The standard deviation, sj${s}_{\mathit{j}}$, of the variable specified in kvar(j)${\mathbf{kvar}}\left(\mathit{j}\right)$, for j = 1,2,,p$\mathit{j}=1,2,\dots ,p$.
3:     sspz(ldsspz,nvars) – double array
ldsspznvars$\mathit{ldsspz}\ge {\mathbf{nvars}}$.
sspz(j,k)${\mathbf{sspz}}\left(\mathit{j},\mathit{k}\right)$ is the cross-product about zero, jk${\stackrel{~}{S}}_{\mathit{j}\mathit{k}}$, for the variables specified in kvar(j)${\mathbf{kvar}}\left(\mathit{j}\right)$ and kvar(k)${\mathbf{kvar}}\left(\mathit{k}\right)$, for j = 1,2,,p$\mathit{j}=1,2,\dots ,p$ and k = 1,2,,p$\mathit{k}=1,2,\dots ,p$.
4:     rz(ldrz,nvars) – double array
ldrznvars$\mathit{ldrz}\ge {\mathbf{nvars}}$.
rz(j,k)${\mathbf{rz}}\left(\mathit{j},\mathit{k}\right)$ is the correlation-like coefficient, jk${\stackrel{~}{R}}_{\mathit{j}\mathit{k}}$, between the variables specified in kvar(j)${\mathbf{kvar}}\left(\mathit{j}\right)$ and kvar(k)${\mathbf{kvar}}\left(\mathit{k}\right)$, for j = 1,2,,p$\mathit{j}=1,2,\dots ,p$ and k = 1,2,,p$\mathit{k}=1,2,\dots ,p$.
5:     ncases – int64int32nag_int scalar
The minimum number of cases used in the calculation of any of the sums of squares and cross-products and correlation-like coefficients (when cases involving missing values have been eliminated).
6:     cnt(ldcnt,nvars) – double array
ldcntnvars$\mathit{ldcnt}\ge {\mathbf{nvars}}$.
cnt(j,k)${\mathbf{cnt}}\left(\mathit{j},\mathit{k}\right)$ is the number of cases, cjk${c}_{\mathit{j}\mathit{k}}$, actually used in the calculation of the sum of cross-product and correlation-like coefficient for the variables specified in kvar(j)${\mathbf{kvar}}\left(\mathit{j}\right)$ and kvar(k)${\mathbf{kvar}}\left(\mathit{k}\right)$, for j = 1,2,,p$\mathit{j}=1,2,\dots ,p$ and k = 1,2,,p$\mathit{k}=1,2,\dots ,p$.
7:     ifail – int64int32nag_int scalar
${\mathrm{ifail}}={\mathbf{0}}$ unless the function detects an error (see [Error Indicators and Warnings]).

## Error Indicators and Warnings

Note: nag_correg_coeffs_zero_subset_miss_pair (g02bm) may return useful information for one or more of the following detected errors or warnings.
Errors or warnings detected by the function:

Cases prefixed with W are classified as warnings and do not generate an error of type NAG:error_n. See nag_issue_warnings.

ifail = 1${\mathbf{ifail}}=1$
 On entry, n < 2${\mathbf{n}}<2$.
ifail = 2${\mathbf{ifail}}=2$
 On entry, nvars < 2${\mathbf{nvars}}<2$, or ${\mathbf{nvars}}>{\mathbf{m}}$.
ifail = 3${\mathbf{ifail}}=3$
 On entry, ldx < n$\mathit{ldx}<{\mathbf{n}}$, or ldsspz < nvars$\mathit{ldsspz}<{\mathbf{nvars}}$, or ldrz < nvars$\mathit{ldrz}<{\mathbf{nvars}}$, or ldcnt < nvars$\mathit{ldcnt}<{\mathbf{nvars}}$.
ifail = 4${\mathbf{ifail}}=4$
 On entry, kvar(j) < 1${\mathbf{kvar}}\left(j\right)<1$, or kvar(j) > m${\mathbf{kvar}}\left(j\right)>{\mathbf{m}}$ for some j = 1,2, … ,nvars$j=1,2,\dots ,{\mathbf{nvars}}$.
W ifail = 5${\mathbf{ifail}}=5$
After observations with missing values were omitted, fewer than two cases remained for at least one pair of variables. (The pairs of variables involved can be determined by examination of the contents of the array cnt.) All means, standard deviations, sums of squares and cross-products, and correlation-like coefficients based on two or more cases are returned by the function even if ${\mathbf{ifail}}={\mathbf{5}}$.

## Accuracy

nag_correg_coeffs_zero_subset_miss_pair (g02bm) does not use additional precision arithmetic for the accumulation of scalar products, so there may be a loss of significant figures for large n$n$.
You are warned of the need to exercise extreme care in your selection of missing values. nag_correg_coeffs_zero_subset_miss_pair (g02bm) treats all values in the inclusive range (1 ± 0.1(x02be2)) × xmj$\left(1±{0.1}^{\left(\mathbf{x02be}-2\right)}\right)×{xm}_{j}$, where xmj${\mathit{xm}}_{j}$ is the missing value for variable j$j$ 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.

The time taken by nag_correg_coeffs_zero_subset_miss_pair (g02bm) depends on n$n$ and p$p$, and the occurrence of missing values.
The function uses a two-pass algorithm.

## Example

```function nag_correg_coeffs_zero_subset_miss_pair_example
x = [3, 3, 1, 2;
6, 4, -1, 4;
9, 0, 5, 9;
12, 2, 0, 0;
-1, 5, 4, 12];
miss = [int64(1);1;0;1];
xmiss = [-1;
0;
0;
0];
kvar = [int64(4);1;2];
[xbar, std, sspz, rz, ncases, count, ifail] = ...
nag_correg_coeffs_zero_subset_miss_pair(x, miss, xmiss, kvar)
```
```

xbar =

6.7500
7.5000
3.5000

std =

4.5735
3.8730
1.2910

sspz =

245   111    82
111   270    57
82    57    54

rz =

1.0000    0.9840    0.9055
0.9840    1.0000    0.7699
0.9055    0.7699    1.0000

ncases =

3

count =

4     3     3
3     4     3
3     3     4

ifail =

0

```
```function g02bm_example
x = [3, 3, 1, 2;
6, 4, -1, 4;
9, 0, 5, 9;
12, 2, 0, 0;
-1, 5, 4, 12];
miss = [int64(1);1;0;1];
xmiss = [-1;
0;
0;
0];
kvar = [int64(4);1;2];
[xbar, std, sspz, rz, ncases, count, ifail] = g02bm(x, miss, xmiss, kvar)
```
```

xbar =

6.7500
7.5000
3.5000

std =

4.5735
3.8730
1.2910

sspz =

245   111    82
111   270    57
82    57    54

rz =

1.0000    0.9840    0.9055
0.9840    1.0000    0.7699
0.9055    0.7699    1.0000

ncases =

3

count =

4     3     3
3     4     3
3     3     4

ifail =

0

```