df_netKmax6: An example of a row-dependent dataset with known network of...

In tmlenet: Targeted Maximum Likelihood Estimation for Network Data

Description

Simulated dataset containing 3 measured i.i.d. baseline covariates (W1, W2, W3), dependent binary exposure (A) and dependent binary binary outcome (Y), along with a known network of friends encoded by strings on space separated friend IDs in Net_str. The baseline covariates (W1,W2,W3) were sampled as i.i.d., while the exposure value of A for each observation i was sampled conditionally on the values of i's baseline covariates (W1[i] W2[i], W3[i]), as well as the baseline covariate values of i's friends in Net_str. Similarly, the binary outcome Y for each observation was generated conditionally on i's exposure and baseline covariates values in (W1[i],W2[i],W3[i],A[i]), as well as the values of exposures and baseline covariates of i's friends in Net_str. Individual variables are described below.

Usage

data(df_netKmax6)

Format

A data frame with 1,000 dependent observations (rows) and 6 variables:

IDs

unique observation identifier

W1

categorical baseline covariate (independent), range 0-5

W2

binary baseline covariate (independent)

W3

binary baseline covariate (independent)

A

binary exposure that depends on unit's baseline covariate values, as well as the baseline covariate values of observations in the friend network Net_str

Y

binary outcome that depends on unit's baseline covariate value and exposure, as well as the baseline covariate values and exposures of observations in the friend network Net_str

nFriends

number of friends for each observation (row), range 0-6

Net_str

a vector of strings, where for each observation its a string of space separated friend IDs (this can be either observation IDs or just space separated friend row numbers)

tmlenet documentation built on May 29, 2017, 2:22 p.m.