,
Takahide Yanagi [aut, cre, cph]
| Title: | Inference on LATEs under Network Interference of Unknown Form |
|---|---|
| Description: | Estimating causal parameters in the presence of treatment spillover is of great interest in statistics. This package provides tools for instrumental variables estimation of average causal effects under network interference of unknown form. The target parameters are the local average direct effect, the local average indirect effect, the local average overall effect, and the local average spillover effect. The methods are developed by Hoshino and Yanagi (2023) <doi:10.48550/arXiv.2108.07455>. |
| Authors: | Tadao Hoshino [aut, cph] ,
Takahide Yanagi [aut, cre, cph]
|
| Maintainer: | Takahide Yanagi <[email protected]> |
| License: | MIT + file LICENSE |
| Version: | 1.0.1 |
| Built: | 2024-11-10 03:44:27 UTC |
| Source: | https://github.com/tkhdyanagi/latenetwork |
The datageneration() function generates artificial ring-network data
by simulation.
The function is used in the package vignette.
datageneration(n)datageneration(n)
n |
The sample size |
A list containing the outcome vector, the treatment vector, the instrumental vector, and the true instrumental exposure vector, and the symmetric binary adjacency matrix.
latenetwork::datageneration(n = 2000)latenetwork::datageneration(n = 2000)
Inference on the average direct effect of the IV on the outcome, that on the treatment receipt, and the local average direct effect in the presence of network spillover of unknown form
direct( Y, D, Z, IEM = NULL, S, A, K = 1, t = NULL, bw = NULL, B = NULL, alp = 0.05 )direct( Y, D, Z, IEM = NULL, S, A, K = 1, t = NULL, bw = NULL, B = NULL, alp = 0.05 )
Y |
An n-dimensional outcome vector |
D |
An n-dimensional binary treatment vector |
Z |
An n-dimensional binary instrumental vector |
IEM |
An n-dimensional instrumental exposure vector.
If |
S |
An n-dimensional logical vector to indicate whether each unit belongs to the sub-population S |
A |
An n times n symmetric binary adjacency matrix |
K |
A scalar to indicate the range of neighborhood
used for constructing the interference set.
Default is 1.
In the |
t |
A scalar of the evaluation point of IEM. Default is NULL. |
bw |
A scalar of the bandwidth used for the HAC estimation and
the wild bootstrap.
If |
B |
The number of bootstrap repetitions.
If |
alp |
The significance level. Default is 0.05. |
The direct() function estimates the average direct effect of the IV
on the outcome, that on the treatment receipt, and
the local average direct effect via inverse probability weighting
in the approximate neighborhood interference framework.
The function also computes the standard errors and the confidence intervals
for the target parameters based on the network HAC estimation and
the wild bootstrap.
For more details, see Hoshino and Yanagi (2023).
The lengths of Y, D, Z, S and
of the row and column of A must be the same.
IEM must be NULL or a vector of the same length as Y.
t must be NULL or a value in the support of IEM.
K must be a positive integer.
bw must be NULL or a non-negative integer.
B must be NULL or a positive number.
alp must be a positive number between 0 and 0.5.
A data.frame containing the following elements:
est |
The parameter estimate |
HAC_SE |
The standard error computed by the network HAC estimation |
HAC_CI_L |
The lower bound of the confidence interval computed by the network HAC estimation |
HAC_CI_U |
The upper bound of the confidence interval computed by the network HAC estimation |
wild_SE |
The standard error computed by the wild bootstrap |
wild_CI_L |
The lower bound of the confidence interval computed by the wild bootstrap |
wild_CI_U |
The upper bound of the confidence interval computed by the wild bootstrap |
bw |
The bandwidth used for the HAC estimation and the wild bootstrap |
size |
The size of the subpopulation S |
Hoshino, T., & Yanagi, T. (2023). Causal inference with noncompliance and unknown interference. arXiv preprint arXiv:2108.07455.
Leung, M.P. (2022). Causal inference under approximate neighborhood interference. Econometrica, 90(1), pp.267-293.
# Generate artificial data set.seed(1) n <- 2000 data <- latenetwork::datageneration(n = n) # Arguments Y <- data$Y D <- data$D Z <- data$Z IEM <- data$IEM S <- rep(TRUE, n) A <- data$A K <- 1 t <- 0 bw <- NULL B <- NULL alp <- 0.05 # Estimation latenetwork::direct(Y = Y, D = D, Z = Z, IEM = IEM, S = S, A = A, K = K, t = t, bw = bw, B = B, alp = alp)# Generate artificial data set.seed(1) n <- 2000 data <- latenetwork::datageneration(n = n) # Arguments Y <- data$Y D <- data$D Z <- data$Z IEM <- data$IEM S <- rep(TRUE, n) A <- data$A K <- 1 t <- 0 bw <- NULL B <- NULL alp <- 0.05 # Estimation latenetwork::direct(Y = Y, D = D, Z = Z, IEM = IEM, S = S, A = A, K = K, t = t, bw = bw, B = B, alp = alp)
Inference on the average indirect effect of the IV on the outcome, that on the treatment receipt, and the local average indirect effect in the presence of network spillover of unknown form
indirect(Y, D, Z, S, A, K = 1, bw = NULL, B = NULL, alp = 0.05)indirect(Y, D, Z, S, A, K = 1, bw = NULL, B = NULL, alp = 0.05)
Y |
An n-dimensional outcome vector |
D |
An n-dimensional binary treatment vector |
Z |
An n-dimensional binary instrumental vector |
S |
An n-dimensional logical vector to indicate whether each unit belongs to the sub-population S |
A |
An n times n symmetric binary adjacency matrix |
K |
A scalar to indicate the range of neighborhood used for constructing the interference set. Default is 1. |
bw |
A scalar of the bandwidth used for the HAC estimation and
the wild bootstrap.
If |
B |
The number of bootstrap repetitions.
If |
alp |
The significance level. Default is 0.05. |
The indirect() function estimates the average indirect effect of the IV
on the outcome, that on the treatment receipt, and
the local average indirect effect via inverse probability weighting
in the approximate neighborhood interference framework.
The function also computes the standard errors and the confidence intervals
for the target parameters based on the network HAC estimation and
the wild bootstrap.
For more details, see Hoshino and Yanagi (2023).
The lengths of Y, D, Z, S and
of the row and column of A must be the same.
K must be a positive integer.
bw must be NULL or a non-negative number.
B must be NULL or a positive number.
alp must be a positive number between 0 and 0.5.
A data.frame containing the following elements:
est |
The parameter estimate |
HAC_SE |
The standard error computed by the network HAC estimation |
HAC_CI_L |
The lower bound of the confidence interval computed by the network HAC estimation |
HAC_CI_U |
The upper bound of the confidence interval computed by the network HAC estimation |
wild_SE |
The standard error computed by the wild bootstrap |
wild_CI_L |
The lower bound of the confidence interval computed by the wild bootstrap |
wild_CI_U |
The upper bound of the confidence interval computed by the wild bootstrap |
bw |
The bandwidth used for the HAC estimation and the wild bootstrap |
size |
The size of the subpopulation S |
Hoshino, T., & Yanagi, T. (2023). Causal inference with noncompliance and unknown interference. arXiv preprint arXiv:2108.07455.
Leung, M.P. (2022). Causal inference under approximate neighborhood interference. Econometrica, 90(1), pp.267-293.
# Generate artificial data set.seed(1) n <- 2000 data <- latenetwork::datageneration(n = n) # Arguments Y <- data$Y D <- data$D Z <- data$Z S <- rep(TRUE, n) A <- data$A K <- 1 bw <- NULL B <- NULL alp <- 0.05 # Estimation latenetwork::indirect(Y = Y, D = D, Z = Z, S = S, A = A, K = K, bw = bw, B = B, alp = alp)# Generate artificial data set.seed(1) n <- 2000 data <- latenetwork::datageneration(n = n) # Arguments Y <- data$Y D <- data$D Z <- data$Z S <- rep(TRUE, n) A <- data$A K <- 1 bw <- NULL B <- NULL alp <- 0.05 # Estimation latenetwork::indirect(Y = Y, D = D, Z = Z, S = S, A = A, K = K, bw = bw, B = B, alp = alp)
Inference on the average overall effect of the IV on the outcome, that on the treatment receipt, and the local average overall effect in the presence of network spillover of unknown form
overall(Y, D, Z, S, A, K = 1, bw = NULL, B = NULL, alp = 0.05)overall(Y, D, Z, S, A, K = 1, bw = NULL, B = NULL, alp = 0.05)
Y |
An n-dimensional outcome vector |
D |
An n-dimensional binary treatment vector |
Z |
An n-dimensional binary instrumental vector |
S |
An n-dimensional logical vector to indicate whether each unit belongs to the sub-population S |
A |
An n times n symmetric binary adjacency matrix |
K |
A scalar to indicate the range of neighborhood used for constructing the interference set. Default is 1. |
bw |
A scalar of the bandwidth used for the HAC estimation and
the wild bootstrap.
If |
B |
The number of bootstrap repetitions.
If |
alp |
The significance level. Default is 0.05. |
The overall() function estimates the average overall effect of the IV
on the outcome, that on the treatment receipt, and
the local average overall effect via inverse probability weighting
in the approximate neighborhood interference framework.
The function also computes the standard errors and the confidence intervals
for the target parameters based on the network HAC estimation and
the wild bootstrap.
For more details, see Hoshino and Yanagi (2023).
The lengths of Y, D, Z, S and
of the row and column of A must be the same.
K must be a positive integer.
bw must be NULL or a non-negative number.
B must be NULL or a positive number.
alp must be a positive number between 0 and 0.5.
A data.frame containing the following elements:
est |
The parameter estimate |
HAC_SE |
The standard error computed by the network HAC estimation |
HAC_CI_L |
The lower bound of the confidence interval computed by the network HAC estimation |
HAC_CI_U |
The upper bound of the confidence interval computed by the network HAC estimation |
wild_SE |
The standard error computed by the wild bootstrap |
wild_CI_L |
The lower bound of the confidence interval computed by the wild bootstrap |
wild_CI_U |
The upper bound of the confidence interval computed by the wild bootstrap |
bw |
The bandwidth used for the HAC estimation and the wild bootstrap |
size |
The size of the subpopulation S |
Hoshino, T., & Yanagi, T. (2023). Causal inference with noncompliance and unknown interference. arXiv preprint arXiv:2108.07455.
Leung, M.P. (2022). Causal inference under approximate neighborhood interference. Econometrica, 90(1), pp.267-293.
# Generate artificial data set.seed(1) n <- 2000 data <- latenetwork::datageneration(n = n) # Arguments Y <- data$Y D <- data$D Z <- data$Z S <- rep(TRUE, n) A <- data$A K <- 1 bw <- NULL B <- NULL alp <- 0.05 # Estimation latenetwork::overall(Y = Y, D = D, Z = Z, S = S, A = A, K = K, bw = bw, B = B, alp = alp)# Generate artificial data set.seed(1) n <- 2000 data <- latenetwork::datageneration(n = n) # Arguments Y <- data$Y D <- data$D Z <- data$Z S <- rep(TRUE, n) A <- data$A K <- 1 bw <- NULL B <- NULL alp <- 0.05 # Estimation latenetwork::overall(Y = Y, D = D, Z = Z, S = S, A = A, K = K, bw = bw, B = B, alp = alp)
Inference on the average spillover effect of the IV on the outcome, that on the treatment receipt, and the local average spillover effect in the presence of network spillover of unknown form
spillover( Y, D, Z, IEM, S, A, K = 1, z, t0, t1, bw = NULL, B = NULL, alp = 0.05 )spillover( Y, D, Z, IEM, S, A, K = 1, z, t0, t1, bw = NULL, B = NULL, alp = 0.05 )
Y |
An n-dimensional outcome vector |
D |
An n-dimensional binary treatment vector |
Z |
An n-dimensional binary instrumental vector |
IEM |
An n-dimensional instrumental exposure vector |
S |
An n-dimensional logical vector to indicate whether each unit belongs to the sub-population S |
A |
An n times n symmetric binary adjacency matrix |
K |
A scalar to indicate the range of neighborhood
used for constructing the interference set.
Default is 1.
In the |
z |
A scalar of the evaluation point of Z |
t0 |
A scalar of the evaluation point of instrumental exposure (from) |
t1 |
A scalar of the evaluation point of instrumental exposure (to) |
bw |
A scalar of the bandwidth used for the HAC estimation and
the wild bootstrap.
If |
B |
The number of bootstrap repetitions.
If |
alp |
The significance level. Default is 0.05. |
The spillover() function estimates the average spillover effect of the IV
on the outcome, that on the treatment receipt,
and the local average spillover effect via inverse probability weighting
in the approximate neighborhood interference framework.
The function also computes the standard errors and the confidence intervals
for the target parameters based on the network HAC estimation and
the wild bootstrap.
For more details, see Hoshino and Yanagi (2023).
The lengths of Y, D, Z, IEM, S and
of the row and column of A must be the same.
z must be 0 or 1.
t0 and t1 must be values in the support of IEM.
bw must be NULL or a non-negative number.
B must be NULL or a positive integer.
alp must be a positive number between 0 and 0.5.
A data frame containing the following elements:
est |
The parameter estimate |
HAC_SE |
The standard error computed by the network HAC estimation |
HAC_CI_L |
The lower bound of the confidence interval computed by the network HAC estimation |
HAC_CI_U |
The upper bound of the confidence interval computed by the network HAC estimation |
wild_SE |
The standard error computed by the wild bootstrap |
wild_CI_L |
The lower bound of the confidence interval computed by the wild bootstrap |
wild_CI_U |
The upper bound of the confidence interval computed by the wild bootstrap |
bw |
The bandwidth used for the HAC estimation and the wild bootstrap |
size |
The size of the subpopulation S |
Hoshino, T., & Yanagi, T. (2023). Causal inference with noncompliance and unknown interference. arXiv preprint arXiv:2108.07455.
Leung, M.P. (2022). Causal inference under approximate neighborhood interference. Econometrica, 90(1), pp.267-293.
# Generate artificial data set.seed(1) n <- 2000 data <- latenetwork::datageneration(n = n) # Arguments Y <- data$Y D <- data$D Z <- data$Z S <- rep(TRUE, n) A <- data$A K <- 1 IEM <- ifelse(A %*% Z > 0, 1, 0) z <- 1 t0 <- 0 t1 <- 1 bw <- NULL B <- NULL alp <- 0.05 # Estimation latenetwork::spillover(Y = Y, D = D, Z = Z, IEM = IEM, S = S, A = A, K = K, z = z, t0 = t0, t1 = t1, bw = bw, B = B, alp = alp)# Generate artificial data set.seed(1) n <- 2000 data <- latenetwork::datageneration(n = n) # Arguments Y <- data$Y D <- data$D Z <- data$Z S <- rep(TRUE, n) A <- data$A K <- 1 IEM <- ifelse(A %*% Z > 0, 1, 0) z <- 1 t0 <- 0 t1 <- 1 bw <- NULL B <- NULL alp <- 0.05 # Estimation latenetwork::spillover(Y = Y, D = D, Z = Z, IEM = IEM, S = S, A = A, K = K, z = z, t0 = t0, t1 = t1, bw = bw, B = B, alp = alp)