Issue with strata in g__lineplot()

[qiliu1013]

g_lineplot() is re-setting the factor level in strata.

adsl <- scda::synthetic_cdisc_data("latest")$adsl
adlb <- scda::synthetic_cdisc_data("latest")$adlb
adlb <- dplyr::filter(adlb, ANL01FL == "Y", PARAMCD == "ALT", AVISIT != "SCREENING")
adlb$AVISIT <- droplevels(adlb$AVISIT)
adlb <- dplyr::mutate(adlb, AVISIT = forcats::fct_reorder(AVISIT, AVISITN, min))
unique(adlb$ARM)
[1] A: Drug X      C: Combination B: Placebo    
Levels: A: Drug X B: Placebo C: Combination

adlb <- adlb %>% mutate(ARM1 = factor(ARM, levels = c("B: Placebo", "C: Combination", "A: Drug X")))
unique(adlb$ARM1)
[1] A: Drug X      C: Combination B: Placebo    
Levels: B: Placebo C: Combination A: Drug X

 tern::g_lineplot(
     adlb, strata = ARM,
     whiskers = "mean_ci_upr",
     title = "Plot of Mean and Upper 95% Confidence Limit by Visit"
 )

tern::g_lineplot(
     adlb, strata = ARM1,
     whiskers = "mean_ci_upr",
     title = "Plot of Mean and Upper 95% Confidence Limit by Visit"
 )

The level of ARM and ARM1 are different, but the outputs are the same.

[qiliu1013]

Hi @shajoezhu, I tested it using the code from help doc. Same issue occurs.

Thanks,
Qi

[edelarua]

Hi @qiliu1013,

g_lineplot uses a control function control_lineplot_vars to customize arguments such as strata for g_lineplot. Since strata is not directly an argument for g_lineplot, differing the variables for strata in your code above do not change the plot. To create these plots, the correct code is:

tern::g_lineplot(
  adlb, 
  variables = control_lineplot_vars(strata = "ARM"),
  whiskers = "mean_ci_upr",
  title = "Plot of Mean and Upper 95% Confidence Limit by Visit"
)

tern::g_lineplot(
  adlb, 
  variables = control_lineplot_vars(strata = "ARM1"),
  whiskers = "mean_ci_upr",
  title = "Plot of Mean and Upper 95% Confidence Limit by Visit"
)

Let me know if you have any questions!