shared3p_statistics_pca.sc
shared3p_statistics_pca.sc
Module for performing principal component analysis.
Functions:
constants
Constants used to specify which result values are needed from the analysis.
Constants
|
|
|
|
|
|
|
|
|
|
|
PCA_RETURN_RESIDUAL - residual matrix. |
|
PCA_RETURN_LOADS - loads. The columns are eigenvectors of the covariance matrix. Can be used to project data to the principal component space. |
|
PCA_RETURN_SCORES - transformed input values. |
|
PCA_RETURN_VARIANCES - variances of principal components. |
|
PCA_RETURN_PROPORTIONS - proportion of variance explained by principal component. |
gspca
Principal component analysis. Note that this method is relatively efficient and precise when a low number of components is required. It uses fixed point numbers internally so it may fail on large inputs due to overflow.
Detailed Description
|
D - shared3p protection domain |
|
Supported types - float32 / float64 |
Parameters
X |
- data matrix where the columns are variables |
n_components |
- how many components to compute |
iterations |
- how many iterations to run the algorithm |
returnValues |
- indicates which results to return. Use bitwise or if you want multiple results. |
|
PCAResult structure |
|
None |
PCAResult
Structure containing the results of the analysis. Note that each field only has a reasonable value if it was requested when calling gspca. See also return value constants.