Skip to contents

Declare diagnosands

Usage

diagnosand_handler(data, ..., subset = NULL, alpha = 0.05, label)

declare_diagnosands(..., handler = diagnosand_handler, label = NULL)

Arguments

data

A data.frame.

...

A set of new diagnosands.

subset

A subset of the simulations data frame within which to calculate diagnosands e.g. subset = p.value < .05.

alpha

Alpha significance level. Defaults to .05.

label

Label for the set of diagnosands.

handler

a tidy-in, tidy-out function

Value

a function that returns a data.frame

Details

If term is TRUE, the names of ... will be returned in a term column, and inquiry will contain the step label. This can be used as an additional dimension for use in diagnosis.

Diagnosands summarize the simulations generated by diagnose_design or simulate_design. Typically, the columns of the resulting simulations data.frame include the following variables: estimate, std.error, p.value, conf.low, conf.high, and inquiry. Many diagnosands will be a function of these variables.

Examples


# Two-arm randomized experiment
design <-
  declare_model(
    N = 500,
    gender = rbinom(N, 1, 0.5),
    X = rep(c(0, 1), each = N / 2),
    U = rnorm(N, sd = 0.25),
    potential_outcomes(Y ~ 0.2 * Z + X + U)
  ) +
  declare_inquiry(ATE = mean(Y_Z_1 - Y_Z_0)) +
  declare_sampling(S = complete_rs(N = N, n = 200)) +
  declare_assignment(Z = complete_ra(N = N, m = 100)) +
  declare_measurement(Y = reveal_outcomes(Y ~ Z)) +
  declare_estimator(Y ~ Z, inquiry = "ATE")

if (FALSE) {
# using built-in defaults:
diagnosis <- diagnose_design(design)
diagnosis

# You can choose your own diagnosands instead of the defaults:

my_diagnosands <-
  declare_diagnosands(median_bias = median(estimate - estimand))

## You can set diagnosands within the diagnose_design function
## using the 'diagnosands =' argument
diagnosis <- diagnose_design(design, diagnosands = my_diagnosands)
diagnosis

## You can also set diagnosands with set_diagnosands 
design <- set_diagnosands(design, diagnosands = my_diagnosands)
diagnosis <- diagnose_design(design)
diagnosis

# If you do not specify diagnosands in diagnose_design, 
#   the function default_diagnosands() is used, 
#   which is reproduced below.

alpha <- 0.05

default_diagnosands <- 
  declare_diagnosands(
    mean_estimand = mean(estimand),
    mean_estimate = mean(estimate),
    bias = mean(estimate - estimand),
    sd_estimate = sqrt(pop.var(estimate)),
    rmse = sqrt(mean((estimate - estimand) ^ 2)),
    power = mean(p.value <= alpha),
    coverage = mean(estimand <= conf.high & estimand >= conf.low)
  )
  
diagnose_design(
  design,
  diagnosands = default_diagnosands
)

# A longer list of potentially useful diagnosands might include:

extended_diagnosands <- 
  declare_diagnosands(
    mean_estimand = mean(estimand),
    mean_estimate = mean(estimate),
    bias = mean(estimate - estimand),
    sd_estimate = sd(estimate),
    rmse = sqrt(mean((estimate - estimand) ^ 2)),
    power = mean(p.value <= alpha),
    coverage = mean(estimand <= conf.high & estimand >= conf.low),
    mean_se = mean(std.error),
    type_s_rate = mean((sign(estimate) != sign(estimand))[p.value <= alpha]),
    exaggeration_ratio = mean((estimate/estimand)[p.value <= alpha]),
    var_estimate = pop.var(estimate),
    mean_var_hat = mean(std.error^2),
    prop_pos_sig = mean(estimate > 0 & p.value <= alpha),
    mean_ci_length = mean(conf.high - conf.low)
  )
  
diagnose_design(
  design,
  diagnosands = extended_diagnosands
)
}