emery (version 0.6.0)

plot_ML: Create plots visualizing the ML estimation process and results.

Description

plot_ML() is a general function for visualizing results generated by estimate_ML().

Usage

plot_ML(ML_est, params = NULL)
plot_ML_binary(
  ML_est,
  params = list(prev = NULL, se = NULL, sp = NULL, D = NULL)
)
plot_ML_ordinal(
  ML_est,
  params = list(pi_1_1 = NULL, phi_1ij_1 = NULL, phi_0ij_1 = NULL, D = NULL)
)
plot_ML_continuous(
  ML_est,
  params = list(prev_1 = NULL, mu_i1_1 = NULL, sigma_i1_1 = NULL, mu_i0_1 = NULL,
    sigma_i0_1 = NULL, D = NULL)
)

Value

A list of ggplot2 plots.

Binary:

prev

A plot showing how the prevalence estimate changes with each iteration of the EM algorithm

se

A plot showing how the sensitivity estimates of each method change with each iteration of the EM algorithm

sp

A plot showing how the specificity estimates of each method change with each iteration of the EM algorithm

qk

A plot showing how the q values for each observation k change over each iteration of the EM algorithm

qk_hist

A histogram of q values. Observations, k, can be colored by True state if it is passed by params$D.

se_sp

A plot showing the path the sensitivity and specificity estimates for each method follows during the EM algorithm. True sensitivity and specificity values can be passed by params$se and params$sp, respectively. This is useful for comparing algorithm results when applied to simulation data where True parameter values are known.

Ordinal:

ROC

The Receiver Operator Characteristic (ROC) curves estimated for each method

q_k1

A plot showing how the q values for each observation, k, change when d=1 over each iteration of the EM algorithm. Observations can be colored by True state if it is passed (params$D).

q_k0

A plot showing how the q values for each observation, k, change when d=0 over each iteration of the EM algorithm. Observations can be colored by True state if it is passed by params$D.

q_k1_hist

A histogram of q_1 values. Observations, k, can be colored by True state if it is passed by params$D.

phi_d

A stacked bar graph representing the estimated CMFs of each method when d=0 and d=1.

Continuous:

ROC

The Receiver Operator Characteristic (ROC) curves estimated for each method

z_k1

A plot showing how the z_k1 values for each observation change over each iteration of the EM algorithm. Observations can be colored by True state if it is passed (params$D).

z_k0

A plot showing how the z_k0 values for each observation change over each iteration of the EM algorithm. Observations can be colored by True state if it is passed (params$D).

z_k1_hist

A histogram of z_k1 values. Observations can be colored by True state if it is passed (params$D).

Arguments

ML_est

A MultiMethodMLEstimate class object

params

A list of population parameters. This is primarily used to evaluate results from a simulation where the target parameters are known, but can be used to visualize results with respect to some True value.

Examples

# Set seed for this example
set.seed(11001101)

# Generate data for 4 binary methods
my_sim <- generate_multimethod_data(
  "binary",
  n_obs = 75,
  n_method = 4,
  se = c(0.87, 0.92, 0.79, 0.95),
  sp = c(0.85, 0.93, 0.94, 0.80),
  method_names = c("alpha", "beta", "gamma", "delta"))

# View the data
my_sim$generated_data

# View the parameters used to generate the data
my_sim$params

# Estimate ML accuracy values by EM algorithm
my_result <- estimate_ML(
  "binary",
  data = my_sim$generated_data,
  save_progress = FALSE # this reduces the data stored in the resulting object
)

# View results of ML estimate
my_result@results