tp_apc: Tall-Project Imputation of Missing Value in Panel Data

In cykbennie/fbi: Factor-Based Imputation and FRED-MD/QD Data Set

Tall-Project Imputation of Missing Value in Panel Data

Description

tp_apc imputes the missing values in a given panel data using the method of "Tall-Project".

Usage

tp_apc(X, kmax, center = FALSE, standardize = FALSE, re_estimate = TRUE)

Arguments

X	a matrix of size T by N with missing values.
kmax	integer, indicating the maximum number of factors.
center	logical, indicating whether or not X should be demeaned
standardize	logical, indicating whether or not X should be scaled.
re_estimate	logical, indicating whether or not output factors, 'Fhat', 'Lamhat', 'Dhat', and 'Chat', should be re-estimated from the imputed data.

Value

a list of elements:

Fhat	estimated F
Lamhat	estimated Lambda
Dhat	estimated D
Chat	euqals Fhat x Lamhat'
ehat	equals Xhat - Chat
data	X with missing data imputed
X	the original data with missing values
kmax	the maximum number of factors
center	logical, indicating whether or not X was demeaned in the algorithm
standardize	logical, indicating whether or not X was scaled in the algorithm
re_estimate	logical, indicating whether or not output factors, 'Fhat', 'Lamhat', 'Dhat', and 'Chat', were re-estimated from the imputed data

Author(s)

Yankang (Bennie) Chen <[e-mail]>

Serena Ng <[e-mail]>

Jushan Bai <[e-mail]>

References

Ercument Cahan, Jushan Bai, and Serena Ng (2021), Factor-Based Imputation of Missing Values and Covariances in Panel Data of Large Dimensions. https://arxiv.org/abs/2103.03045

cykbennie/fbi documentation built on Jan. 24, 2025, 5:59 p.m.