factorPull: factorPull
In NickCH-K/MagnifiedIV: Magnified Instrumental Variables Estimator

Description Usage Arguments Details Examples

View source: R/factorPull.R

This function takes any estimated regression compatible with broom::tidy() that contains a factor variable, or a factor variable interacted with something else. It will pull out those coefficients and line them up with the original data so you can store them as a variable.

factorPull(
  model,
  data,
  factor,
  interaction = NULL,
  specify = NULL,
  basevalue = 0,
  addterm = NULL,
  value = "estimate",
  na.predict = TRUE
)

`model`	A model object compatible with `broom::tidy()`.
`data`	A data set containing the factor variable you want coefficients for.
`factor`	A string with the name of the factor variable you'd like to get the coefficients of. Note if the variable is included in regression via the `factor()` function, include that. So if your model is `mpg~factor(cyl)+hp`, do `factor = 'factor(cyl)'`.
`interaction`	A string with the name of a variable being interacted with `factor` for which you'd like to get the different interaction terms. Specify the name as it shows up in the regression table. So for example if `factor = 'X'` and your model has coefficients `Z1:X1, Z1:X2, Z2:X1, Z2:X2`, then `interaction = 'Z1'` will get the `Z1:X1, Z1:X2` coefficients.
`specify`	For more complex sets of interactions, like three-way interactions, or if your `factor` variable is included via some function more complex than `factor()`, you can specify the exact naming structure of the coefficients, with `{value}` standing in for the factor value. So for example if you had coefficients `Z:X1:W, Z:X2:W` in your model, you could do `specify = 'Z:X{value}:W'`. This will override anything in `interaction`.
`basevalue`	If `specify` is used, what value should the `factor` reference group (or any excluded terms) be given? By default `0`. Set to `NA` to leave omitted terms as `NA`.
`addterm`	A string indicating a coefficient in the model to be added to all terms. Commonly this is the coefficient for the variable being interacted with, to give an overall effect of that variable, or `'(Intercept)'` to give a mean prediction.
`value`	A string with the column name of the `broom::tidy()` result that you would like to extract. By default this is `value = 'estimate'` to get the coefficients.
`na.predict`	Should observations in the data but not in the model, which do have nonmissing values of `factor`, be included in predictions? Setting this to `FALSE` requires that sending your `model` through `fitted()` produces a named vector.

If multiple coefficients related to the factor are dropped from the model, be sure to check the result.

df <- data.frame(w1 = rnorm(1000),
                 w2 = rnorm(1000),
                 e1 = rnorm(1000),
                 e2 = rnorm(1000),
                 z = rnorm(1000),
                 groups = factor(floor(0:999/100)))
df$x <- df$w1+df$w2+df$z+df$e1

# Create a model with an interaction between z and groups
lm_inx <- lm(x ~ z*groups + w1 + w2, data = df)

# Get the effect of z for each group
indivfx <- factorPull(lm_inx,
                      data = df,
                      factor = 'groups',
                      interaction = 'z',
                      addterm = 'z')

# Create a model with a fixed effect for groups
lm_inx2 <- lm(x ~ groups + z + w1 + w2, data = df)

# Get the fixed effect for each group, with 0 for the reference group
indivfe <- factorPull(lm_inx,
                      data = df,
                      factor = 'groups')
# Or add in the intercept to get the full fixed effect
indivfe2 <- factorPull(lm_inx,
                       data = df,
                       factor = 'groups',
                       addterm = '(Intercept)')