Estimate smoothed exposure-response function (ERF) for pseudo population

Estimate smoothed exposure-response function (ERF) for matched and weighted data set using non-parametric models.

Usage

estimate_npmetric_erf(
  m_Y,
  m_w,
  counter_weight,
  bw_seq,
  w_vals,
  nthread,
  kernel_appr = "locpol"
)

Arguments

m_Y

A vector of outcome variable in the matched set.

m_w

A vector of continuous exposure variable in the matched set.

counter_weight

A vector of counter or weight variable in the matched set.

bw_seq

A vector of bandwidth values.

w_vals

A vector of values that you want to calculate the values of the ERF at.

nthread

The number of available cores.

kernel_appr

Internal kernel approach. Available options are locpol and kernsmooth.

Value

The function returns a gpsm_erf object. The object includes the following attributes:

• params

• m_Y

• m_w

• bw_seq

• w_vals

• erf

• fcall

Details

Estimate Functions Using Local Polynomial kernel regression.