cov |

Compute covariances, correlations, and other measures of association for the columns in a matrix.

You may compute measures related to Pearson product-moment (ordinary) covariances and correlations, Spearman rank covariances, or Kendall’s tau along with test statistics for evaluating whether the correlations are equal to zero.

Syntax

matrix_name.cov(options) [keywords [@partial z1 z2 z3...]]

You should specify keywords indicating the statistics you wish to display from the list below, optionally followed by the keyword @partial and the name of a conditioning matrix. The columns should contain the conditioning variables, and the number of rows should match the original matrix.

You may specify keywords from one of the four sets (Pearson correlation, Spearman correlation, Kendall’s tau, Uncentered Pearson) corresponding the computational method you wish to employ. (You may not select keywords from more than one set.)

If you do not specify keywords, EViews will assume “cov” and compute the Pearson covariance matrix. Note that Matrix::cov is equivalent to the Matrix::cor command with a different default setting.

Pearson Correlation

cov | Product moment covariance. |

corr | Product moment correlation. |

sscp | Sums-of-squared cross-products. |

stat | Test statistic (t-statistic) for evaluating whether the correlation is zero. |

prob | Probability under the null for the test statistic. |

cases | Number of cases. |

obs | Number of observations. |

wgts | Sum of the weights. |

Spearman Rank Correlation

rcov | Spearman’s rank covariance. |

rcorr | Spearman’s rank correlation. |

rsscp | Sums-of-squared cross-products. |

rstat | Test statistic (t-statistic) for evaluating whether the correlation is zero. |

rprob | Probability under the null for the test statistic. |

cases | Number of cases. |

obs | Number of observations. |

wgts | Sum of the weights. |

Kendall’s tau

taub | Kendall’s tau-b. |

taua | Kendall’s tau-a. |

taucd | Kendall’s concordances and discordances. |

taustat | Kendall’s score statistic for evaluating whether the Kendall’s tau-b measure is zero. |

tauprob | Probability under the null for the score statistic. |

cases | Number of cases. |

obs | Number of observations. |

wgts | Sum of the weights. |

Uncentered Pearson

ucov | Product moment covariance. |

ucorr | Product moment correlation. |

usscp | Sums-of-squared cross-products. |

ustat | Test statistic (t-statistic) for evaluating whether the correlation is zero. |

uprob | Probability under the null for the test statistic. |

cases | Number of cases. |

obs | Number of observations. |

wgts | Sum of the weights. |

Note that cases, obs, and wgts are available for each of the methods.

Options

wgt=name (optional) | Name of vector containing weights. The number of rows of the weight vector should match the number of rows in the original matrix. |

wgtmethod=arg (default = “sstdev”) | Weighting method (when weights are specified using “weight=”): frequency (“freq”), inverse of variances (“var”), inverse of standard deviation (“stdev”), scaled inverse of variances (“svar”), scaled inverse of standard deviations (“sstdev”). Only applicable for ordinary (Pearson) calculations. Weights specified by “wgt=” are frequency weights for rank correlation and Kendall’s tau calculations. |

pairwise | Compute using pairwise deletion of observations with missing cases (pairwise samples). |

df | Compute covariances with a degree-of-freedom correction for the mean (for centered specifications), and any partial conditioning variables. |

multi=arg (default=“none”) | Adjustment to p-values for multiple comparisons: none (“none”), Bonferroni (“bonferroni”), Dunn-Sidak (“dunn”). |

outfmt=arg (default= “single”) | Output format: single table (“single”), multiple table (“mult”), list (“list”), spreadsheet (“sheet”). Note that “outfmt=sheet” is only applicable if you specify a single statistic keyword. |

out=name | Basename for saving output. All results will be saved in Sym matrices named using keys (“COV”, “CORR”, “SSCP”, “TAUA”, “TAUB”, “CONC” (Kendall’s concurrences), “DISC” (Kendall’s discordances), “CASES”, “OBS”, “WGTS”) appended to the basename (e.g., the covariance specified by “out=my” is saved in the Sym matrix “MYCOV”). |

prompt | Force the dialog to appear from within a program. |

p | Print the result. |

Examples

mat1.cov

displays a Pearson covariance matrix for the columns series in MAT1.

mat1.cov corr stat prob

displays a table containing the Pearson covariance, t-statistic for testing for zero correlation, and associated p-value, for the columns in MAT1.

mat1.cov(pairwise) taub taustat tauprob

computes the Kendall’s tau-b, score statistic, and p-value for the score statistic, using samples with pairwise missing value exclusion.

mat1.cov(out=aa) cov

computes the Pearson covariance for the columns in MAT1 and saves the results in the symmetric matrix object AACO.

Cross-references