Fit  linear regression model based on the principal components scores for each  patent indicator data object
Extract  functional principal component analysis coefficients for building linear  regression model and determine the variance of coefficients
Combine  the fPCA  coefficients with the corresponding fPCA  harmonic function for each patent indicator data object
Combine  the fPCA  coefficient variances with the square of the corresponding fPCA  harmonic function for each  patent indicator data object