nice_plots: Funciton to plot nice plots
In mshasan/OPWpaper: Reproduce figures and simulations of OPWeight

Description Usage Arguments Details Value Author(s) References Examples

OPWpaper has stored .RDATA from the simulations. This function will use those simulated data to plot

1
2
3

nice_plots(x_vec, y_matrix, fdr = TRUE, power = TRUE,
  low_eff_plot = FALSE, null = NULL, cv = NULL, ey = NULL, cor = NULL,
  figure = c("ranksProb", "nullPropVsPower", "effectVsFPFP", "CV"))

`x_vec`	A numeric vector corresponds to the x-axis
`y_matrix`	A numeric matrix corresponds to the y-axix
`fdr`	A character vector of ("TRUE" or "FALSE"), determine whether the FDR or FWER will be used, default is FDR.
`power`	A character vector of ("TRUE" or "FALSE"), determine whether the power will be plotted, default is TRUE
`low_eff_plot`	A character vector of ("TRUE" or "FALSE"), deteremine whether the power of the low effect sizes will be plotted, default is FALSE
`null`	Numeric, proportion of the true null if power or FDR/FWER is plotted against the effect sizes
`cv`	Numeric, coefficient of variation of the test statistics
`ey`	Numeric, the value of the effect size if power is plotted against the proportion of the true null tests.
`cor`	Numeric, the correlation coefficient if the figure is for the ranks probability
`figure`	A character vector of c("ranksProb", "nullPropVsPower", "effectVsFPFP", "CV"), determine the types of figure will be plotted

OPWeight package proposed methods to compute the ranks probabilities of the covariate given the test effect size to obtian the optimal power. This function is desigend to plot the power curves under different scenerios. Note that, we alreday simulated power and FDR/FWER for the different scenerios and stored in the packages OPWpaper as .RDATA. This function will only be able to use those data sets or data with the similar formats.

A plot of multiple curves

Mohamad S. Hasan, shakilmohamad7@gmail.com

Hasan and Schliekelman (2017)

# examples from the previously stored .RDATA
# plot of power against the effect sizes

load(system.file("simulations/results", "simu_fwerPowerFdrPower_cont.RDATA",
                package = "OPWpaper"), envir = environment())
ey_vec <- c(seq(0, 1, .2), 2, 3, 5, 8)
p_.5_eq_power <- nice_plots(x_vec = ey_vec, y_matrix = FwerPowerFdrPower2e1,
                                null = 50, figure = "effectVsFPFP")

# p_.9_eq_power <- nice_plots(x_vec = ey_vec, y_matrix = FwerPowerFdrPower4e1,
#                                null = 90, figure = "effectVsFPFP")
# p_.99_eq_power<- nice_plots(x_vec = ey_vec, y_matrix = FwerPowerFdrPower5e1,
#                                null = 99, figure = "effectVsFPFP")

# p_.5_low_ef_eq_power <- nice_plots(x_vec = ey_vec, y_matrix = FwerPowerFdrPower2e1,
#                       null = 50, low_eff_plot = TRUE, figure = "effectVsFPFP")
# p_.9_low_ef_eq_power <- nice_plots(x_vec = ey_vec, y_matrix = FwerPowerFdrPower4e1,
#                       null = 90, low_eff_plot = TRUE, figure = "effectVsFPFP")
# p_.99_low_ef_eq_power<- nice_plots(x_vec = ey_vec, y_matrix = FwerPowerFdrPower5e1,
#                       null = 99, low_eff_plot = TRUE, figure = "effectVsFPFP")

# p_eq_power = plot_grid(p_.5_eq_power, p_.9_eq_power, p_.99_eq_power,
#                    p_.5_low_ef_eq_power, p_.9_low_ef_eq_power, p_.99_low_ef_eq_power,
#                    ncol = 3, labels = letters[1:3], align = 'hv')
# title <- ggdraw() + draw_label("Power: et = ey")
# plot_grid(title, p_eq_power, legend, ncol = 1, rel_heights=c(.1, 1, .1))

# plot of power against the true propotion of the null
# mat_ef.6 <- rbind(FwerPowerFdrPower1f1[13:16, 4], FwerPowerFdrPower2f1[13:16, 4],
# FwerPowerFdrPower3f1[13:16, 4], FwerPowerFdrPower4f1[13:16, 4],
# FwerPowerFdrPower5f1[13:16, 4])
# p_ef.6 <- nice_plots(x_vec = nullProp, y_matrix = mat_ef.6, fdr = TRUE,
# power = TRUE, ey = 0.6, figure = "nullPropVsPower")