The function may be called by the names: g07gac, nag_univar_outlier_peirce_1var or nag_outlier_peirce.
g07gac flags outlying values in data using Peirce's criterion. Let
denote a vector of observations (for example the residuals) obtained from a model with parameters,
denote the number of potential outlying values,
and denote the mean and variance of respectively,
denote a vector of length constructed by dropping the values from
with the largest value of ,
denote the (unknown) variance of ,
denote the ratio of and with
Peirce's method flags as a potential outlier if , where and is obtained from the solution of
and is the cumulative distribution function for the standard Normal distribution.
As is unknown an assumption is made that the relationship between and , hence , depends only on the sum of squares of the rejected observations and the ratio estimated as
A value for the cutoff is calculated iteratively. An initial value of is used and a value of is estimated using equation (1). Equation (3) is then used to obtain an estimate of and then equation (2) is used to get a new estimate for . This process is then repeated until the relative change in between consecutive iterations is , where is machine precision.
By construction, the cutoff for testing for potential outliers is less than the cutoff for testing for potential outliers. Therefore, Peirce's criterion is used in sequence with the existence of a single potential outlier being investigated first. If one is found, the existence of two potential outliers is investigated etc.
If one of a duplicate series of observations is flagged as an outlier, then all of them are flagged as outliers.
Gould B A (1855) On Peirce's criterion for the rejection of doubtful observations, with tables for facilitating its application The Astronomical Journal45
Peirce B (1852) Criterion for the rejection of doubtful observations The Astronomical Journal45
1: – IntegerInput
On entry: , the number of observations.
2: – IntegerInput
On entry: , the number of parameters in the model used in obtaining the . If is an observed set of values, as opposed to the residuals from fitting a model with parameters, then should be set to , i.e., as if a model just containing the mean had been used.
3: – const doubleInput
On entry: , the data being tested.
4: – doubleInput
On entry: if , mean must contain , the mean of , otherwise mean is not referenced and the mean is calculated from the data supplied in y.
5: – doubleInput
On entry: if , var must contain , the variance of , otherwise the variance is calculated from the data supplied in y.
6: – IntegerOutput
On exit: the indices of the values in y sorted in descending order of the absolute difference from the mean, therefore,
, for .
7: – Integer *Output
On exit: the number of potential outliers. The indices for these potential outliers are held in the first niout elements of iout. By construction there can be at most values flagged as outliers.
8: – IntegerInput
On entry: the maximum number of values to be returned in arrays diff and llamb.
On exit: if diff is not NULL,
holds for observation , for .
10: – doubleOutput
On exit: if llamb is not NULL,
holds for observation , for .
11: – NagError *Input/Output
The NAG error argument (see Section 7 in the Introduction to the NAG Library CL Interface).
6Error Indicators and Warnings
Dynamic memory allocation failed.
See Section 3.1.2 in the Introduction to the NAG Library CL Interface for further information.
On entry, argument had an illegal value.
On entry, .
On entry, and .
An internal error has occurred in this function. Check the function call and any array sizes. If the call is correct then please contact NAG for assistance.
See Section 7.5 in the Introduction to the NAG Library CL Interface for further information.
Your licence key may have expired or may not have been installed correctly.
See Section 8 in the Introduction to the NAG Library CL Interface for further information.
8Parallelism and Performance
Background information to multithreading can be found in the Multithreading documentation.
g07gac is not threaded in any implementation.
One problem with Peirce's algorithm as implemented in g07gac is the assumed relationship between , the variance using the full dataset, and , the variance with the potential outliers removed. In some cases, for example if the data were the residuals from a linear regression, this assumption may not hold as the regression line may change significantly when outlying values have been dropped resulting in a radically different set of residuals. In such cases g07gbc should be used instead.
This example reads in a series of data and flags any potential outliers.
The dataset used is from Peirce's original paper and consists of fifteen observations on the vertical semidiameter of Venus.