Impute hormones by cycle day

Author

Affiliation

Ruben Arslan

 

Published

Jan. 4, 2021

DOI

 
library(tidyverse)

── Attaching packages ───────────────────────────── tidyverse 1.3.1 ──
✓ ggplot2 3.3.5     ✓ purrr   0.3.4
✓ tibble  3.1.4     ✓ dplyr   1.0.7
✓ tidyr   1.1.3     ✓ stringr 1.4.0
✓ readr   2.0.1     ✓ forcats 0.5.1
── Conflicts ──────────────────────────────── tidyverse_conflicts() ──
x dplyr::filter() masks stats::filter()
x dplyr::lag()    masks stats::lag()
library(codebook)
library(brms)

Loading required package: Rcpp
Loading 'brms' package (version 2.16.2). Useful instructions
can be found by typing help('brms'). A more detailed introduction
to the package is available through vignette('brms_overview').

Attaching package: 'brms'
The following object is masked from 'package:stats':

    ar
plot_curve <- function(model, method = "fitted") {
  cond <- tibble(.rows = 1)
  cond[[names(model$data)[2]]] <- "none"
  plot(conditional_effects(model, method = method,
                conditions = cond, 
                spaghetti = TRUE, nsamples = 100), 
              points = TRUE, 
              point_args = list(alpha = 0.05),
              spaghetti_args = list(alpha = 0.01, colour = '#FF000007'), 
              line_args = list(size = 0), 
              plot = FALSE)[[1]]
}
rmse <- function(error) { sqrt(mean(error^2, na.rm = TRUE)) }
rmse_brms <- function(model) {
  resid <- residuals(model, summary = F)
  rmse_samples <- apply(resid, 1, rmse)
  rmse_ci <- rstantools::posterior_interval(as.matrix(rmse_samples), prob = 0.95)
  sprintf("%.2f [%.2f;%.2f]", mean(rmse_samples), rmse_ci[1], rmse_ci[2]) 
}
mae <- function(error) { mean(abs(error), na.rm = TRUE) }
mae_brms <- function(model) {
  resid <- residuals(model, summary = F)
  mae_samples <- apply(resid, 1, mae)
  mae_ci <- rstantools::posterior_interval(as.matrix(mae_samples), prob = 0.95)
  sprintf("%.2f [%.2f;%.2f]", mean(mae_samples), mae_ci[1], mae_ci[2]) 
}

loo_ci <- function(model) {
  loo_R <- loo_R2(model, re_formula = NA)
  loo_R <- sqrt(loo_R)
  sprintf("%.2f [%.2f;%.2f]", loo_R[,"Estimate"], loo_R[,"Q2.5"], loo_R[,"Q97.5"]) 
}

biocycle <- readRDS("biocycle.rds")
options(mc.cores = parallel::detectCores(), brms.backend = "cmdstanr", brms.file_refit = "on_change")
rstan::rstan_options(auto_write = TRUE)

biocycle <- biocycle %>% filter(between(cycle_length, 20, 35))
biocycle$logE_P <- log(biocycle$E_P)
biocycle$logE <- log(biocycle$total_estradiol)
biocycle$logP <- log(biocycle$progesterone)
biocycle$logT <- log(biocycle$Testo)
biocycle$logFE <- log(biocycle$estradiol)
biocycle$logFT <- log(biocycle$`Free T (ng/dL)`)
biocycle$fertile_lh <- NULL

Days

bc_days <- data.frame(
  bc_day = c(-28:-1, -29:-40),
  prc_stirn_bc = c(.01, .01, .02, .03, .05, .09, .16, .27, .38, .48, .56, .58, .55, .48, .38, .28, .20, .14, .10, .07, .06, .04, .03, .02, .01, .01, .01, .01, rep(.01, times = 12)),
#                   rep(.01, times = 70)), # gangestad uses .01 here, but I think such cases are better thrown than kept, since we might simply have missed a mens
  prc_wcx_bc = c(.000, .000, .001, .002, .004, .009, .018, .032, .050, .069, .085, .094, .093, .085, .073, .059, .047, .036, .028, .021, .016, .013, .010, .008, .007, .006, .005, .005, rep(.005, times = 12))
) %>% arrange(bc_day)

fc_days <- data.frame(
  fc_day = c(0:39),
  prc_stirn_fc = c(.01, .01, .02, .03, .05, .09, .16, .27, .38, .48, .56, .58, .55, .48, .38, .28, .20, .14, .10, .07, .06, .04, .03, .02, .01, .01, .01, .01, rep(.01, times = 12)),
#                   rep(.01, times = 70)), # gangestad uses .01 here, but I think such cases are better thrown than kept, since we might simply have missed a mens
  prc_wcx_fc = c(.000, .000, .001, .002, .004, .009, .018, .032, .050, .069, .085, .094, .093, .085, .073, .059, .047, .036, .028, .021, .016, .013, .010, .008, .007, .006, .005, .005, rep(.005, times = 12))
)

lh_days <- tibble(
    conception_risk_lh = c(rep(0,8), 0.00, 0.01, 0.02, 0.06, 0.16, 0.20, 0.25, 0.24, 0.10, 0.02, 0.02, rep(0,12)),
    lh_day = -15:15
  ) %>% 
    mutate(fertile_lh = conception_risk_lh/max(conception_risk_lh)*0.25/0.31)
biocycle <- biocycle %>% left_join(lh_days %>% select(lh_day, fertile_lh), by = "lh_day")

Impute E

Backward counting

(mod_e_bc = brm(bf(logE | cens(estradiol_cens) ~ s(bc_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle, control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_e_bc'))

Warning: Rows containing NAs were excluded from the model.
 Family: gaussian 
  Links: mu = identity; sigma = identity 
Formula: logE | cens(estradiol_cens) ~ s(bc_day) + (1 | id) + (1 | id:cycle) 
   Data: biocycle (Number of observations: 3425) 
  Draws: 4 chains, each with iter = 1000; warmup = 0; thin = 1;
         total post-warmup draws = 4000

Smooth Terms: 
               Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sds(sbc_day_1)     5.18      1.26     3.29     8.24 1.00      791
               Tail_ESS
sds(sbc_day_1)     1306

Group-Level Effects: 
~id (Number of levels: 246) 
              Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sd(Intercept)     0.29      0.02     0.25     0.32 1.00     1304
              Tail_ESS
sd(Intercept)     2659

~id:cycle (Number of levels: 448) 
              Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sd(Intercept)     0.07      0.03     0.01     0.12 1.01      317
              Tail_ESS
sd(Intercept)      601

Population-Level Effects: 
          Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
Intercept     4.45      0.02     4.41     4.49 1.00     1054     1721
sbc_day_1   -13.26      1.71   -16.66    -9.89 1.00     2509     2890

Family Specific Parameters: 
      Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
sigma     0.49      0.01     0.48     0.51 1.00     3859     3058

Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
and Tail_ESS are effective sample size measures, and Rhat is the potential
scale reduction factor on split chains (at convergence, Rhat = 1).
performance::variance_decomposition(mod_e_bc)

# Random Effect Variances and ICC

Conditioned on: all random effects

## Variance Ratio (comparable to ICC)
Ratio: 0.14  CI 95%: [0.09 0.19]

## Variances of Posterior Predicted Distribution
Conditioned on fixed effects: 0.53  CI 95%: [0.50 0.56]
Conditioned on rand. effects: 0.61  CI 95%: [0.58 0.64]

## Difference in Variances
Difference: 0.08  CI 95%: [0.05 0.12]
# performance::model_performance(mod_e_bc, metrics = c("RMSE", "R2", "R2_adj"))
loo_R2(mod_e_bc, re_formula = NA)

Warning: Results may not be meaningful for censored models.
    Estimate  Est.Error      Q2.5    Q97.5
R2 0.4590614 0.01175674 0.4353065 0.481754
rmse_brms(mod_e_bc)

Warning: Results may not be meaningful for censored models.
[1] "0.49 [0.48;0.49]"
mae_brms(mod_e_bc)

Warning: Results may not be meaningful for censored models.
[1] "0.39 [0.38;0.39]"
plot_curve(mod_e_bc)

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
bc_days[,paste0("est_estradiol_bc", c("", "_se", "_low", "_high"))] <- round(predict(mod_e_bc, newdata = bc_days, re_formula = NA),2)

Forward counting

(mod_e_fc = brm(bf(logE | cens(estradiol_cens) ~ s(fc_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle, control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_e_fc'))

Warning: Rows containing NAs were excluded from the model.
 Family: gaussian 
  Links: mu = identity; sigma = identity 
Formula: logE | cens(estradiol_cens) ~ s(fc_day) + (1 | id) + (1 | id:cycle) 
   Data: biocycle (Number of observations: 3425) 
  Draws: 4 chains, each with iter = 1000; warmup = 0; thin = 1;
         total post-warmup draws = 4000

Smooth Terms: 
               Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sds(sfc_day_1)     2.00      0.61     1.12     3.43 1.00     1260
               Tail_ESS
sds(sfc_day_1)     2183

Group-Level Effects: 
~id (Number of levels: 246) 
              Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sd(Intercept)     0.29      0.02     0.26     0.33 1.00      997
              Tail_ESS
sd(Intercept)     2069

~id:cycle (Number of levels: 448) 
              Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sd(Intercept)     0.03      0.02     0.00     0.08 1.01      398
              Tail_ESS
sd(Intercept)      983

Population-Level Effects: 
          Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
Intercept     4.45      0.02     4.41     4.49 1.00     1192     2141
sfc_day_1    -1.03      1.61    -4.22     2.16 1.00     1753     2502

Family Specific Parameters: 
      Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
sigma     0.57      0.01     0.55     0.58 1.00     4618     2866

Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
and Tail_ESS are effective sample size measures, and Rhat is the potential
scale reduction factor on split chains (at convergence, Rhat = 1).
loo_R2(mod_e_fc, re_formula = NA)

Warning: Results may not be meaningful for censored models.
    Estimate  Est.Error      Q2.5     Q97.5
R2 0.3293852 0.01217784 0.3053314 0.3523772
rmse_brms(mod_e_fc)

Warning: Results may not be meaningful for censored models.
[1] "0.56 [0.56;0.56]"
mae_brms(mod_e_fc)

Warning: Results may not be meaningful for censored models.
[1] "0.45 [0.45;0.46]"
plot_curve(mod_e_fc, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
fc_days[,paste0("est_estradiol_fc", c("", "_se", "_low", "_high"))] <- round(predict(mod_e_fc, newdata = fc_days, re_formula = NA),2)

Relative to LH surge

(mod_e_lh = brm(bf(logE | cens(estradiol_cens) ~ s(lh_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle %>% filter(between(lh_day, -20, 14)), control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_e_lh')
)

Warning: Rows containing NAs were excluded from the model.
 Family: gaussian 
  Links: mu = identity; sigma = identity 
Formula: logE | cens(estradiol_cens) ~ s(lh_day) + (1 | id) + (1 | id:cycle) 
   Data: biocycle %>% filter(between(lh_day, -20, 14)) (Number of observations: 2590) 
  Draws: 4 chains, each with iter = 1000; warmup = 0; thin = 1;
         total post-warmup draws = 4000

Smooth Terms: 
               Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sds(slh_day_1)     8.38      2.01     5.54    13.26 1.00     1026
               Tail_ESS
sds(slh_day_1)     1645

Group-Level Effects: 
~id (Number of levels: 214) 
              Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sd(Intercept)     0.27      0.02     0.24     0.31 1.00     1009
              Tail_ESS
sd(Intercept)     2117

~id:cycle (Number of levels: 340) 
              Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sd(Intercept)     0.07      0.03     0.01     0.12 1.02      288
              Tail_ESS
sd(Intercept)      576

Population-Level Effects: 
          Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
Intercept     4.46      0.02     4.42     4.51 1.00     1180     1664
slh_day_1     0.12      2.06    -3.87     4.27 1.00     2186     2502

Family Specific Parameters: 
      Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
sigma     0.46      0.01     0.45     0.47 1.00     4236     3057

Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
and Tail_ESS are effective sample size measures, and Rhat is the potential
scale reduction factor on split chains (at convergence, Rhat = 1).
loo_R2(mod_e_lh, re_formula = NA)

Warning: Results may not be meaningful for censored models.
    Estimate  Est.Error      Q2.5     Q97.5
R2 0.5182738 0.01413739 0.4900016 0.5460118
rmse_brms(mod_e_lh)

Warning: Results may not be meaningful for censored models.
[1] "0.45 [0.45;0.46]"
mae_brms(mod_e_lh)

Warning: Results may not be meaningful for censored models.
[1] "0.36 [0.35;0.36]"
plot_curve(mod_e_lh, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
lh_days[,paste0("est_estradiol_lh", c("", "_se", "_low", "_high"))] <- round(predict(mod_e_lh, newdata = lh_days, re_formula = NA),2)

Impute FE

Backward counting

(mod_fe_bc = brm(bf(logFE | cens(estradiol_cens) ~ s(bc_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle, control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_fe_bc'))

Warning: Rows containing NAs were excluded from the model.
 Family: gaussian 
  Links: mu = identity; sigma = identity 
Formula: logFE | cens(estradiol_cens) ~ s(bc_day) + (1 | id) + (1 | id:cycle) 
   Data: biocycle (Number of observations: 3425) 
  Draws: 4 chains, each with iter = 1000; warmup = 0; thin = 1;
         total post-warmup draws = 4000

Smooth Terms: 
               Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sds(sbc_day_1)     5.89      1.58     3.70     9.72 1.00      952
               Tail_ESS
sds(sbc_day_1)     1290

Group-Level Effects: 
~id (Number of levels: 246) 
              Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sd(Intercept)     0.28      0.02     0.25     0.32 1.00     1096
              Tail_ESS
sd(Intercept)     2224

~id:cycle (Number of levels: 448) 
              Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
sd(Intercept)     0.08      0.03     0.01     0.14 1.01      343
              Tail_ESS
sd(Intercept)      368

Population-Level Effects: 
          Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
Intercept     0.42      0.02     0.38     0.46 1.00     1624     2302
sbc_day_1   -14.50      1.98   -18.44   -10.56 1.00     2834     2460

Family Specific Parameters: 
      Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
sigma     0.56      0.01     0.54     0.57 1.00     3912     2790

Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
and Tail_ESS are effective sample size measures, and Rhat is the potential
scale reduction factor on split chains (at convergence, Rhat = 1).
loo_R2(mod_fe_bc, re_formula = NA)

Warning: Results may not be meaningful for censored models.
    Estimate  Est.Error      Q2.5     Q97.5
R2 0.4575188 0.01199411 0.4334088 0.4797044
rmse_brms(mod_fe_bc)

Warning: Results may not be meaningful for censored models.
[1] "0.55 [0.54;0.55]"
mae_brms(mod_fe_bc)

Warning: Results may not be meaningful for censored models.
[1] "0.43 [0.43;0.44]"
plot_curve(mod_fe_bc, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
bc_days[,paste0("est_free_estradiol_bc", c("", "_se", "_low", "_high"))] <- round(predict(mod_fe_bc, newdata = bc_days, re_formula = NA),2)

Forward counting

mod_fe_fc = brm(bf(logFE | cens(estradiol_cens) ~ s(fc_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle, control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_fe_fc')

Warning: Rows containing NAs were excluded from the model.
loo_R2(mod_fe_fc, re_formula = NA)

Warning: Results may not be meaningful for censored models.
    Estimate  Est.Error     Q2.5     Q97.5
R2 0.3296142 0.01229912 0.305068 0.3526868
rmse_brms(mod_fe_fc)

Warning: Results may not be meaningful for censored models.
[1] "0.63 [0.63;0.63]"
mae_brms(mod_fe_fc)

Warning: Results may not be meaningful for censored models.
[1] "0.51 [0.50;0.51]"
plot_curve(mod_fe_fc, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
fc_days[,paste0("est_free_estradiol_fc", c("", "_se", "_low", "_high"))] <- round(predict(mod_fe_fc, newdata = fc_days, re_formula = NA),2)

Relative to LH surge

mod_fe_lh = brm(bf(logFE | cens(estradiol_cens) ~ s(lh_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle %>% filter(between(lh_day, -20, 14)), control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_fe_lh')

Warning: Rows containing NAs were excluded from the model.
loo_R2(mod_fe_lh, re_formula = NA)

Warning: Results may not be meaningful for censored models.
    Estimate  Est.Error      Q2.5     Q97.5
R2 0.5213177 0.01417518 0.4931336 0.5481039
rmse_brms(mod_fe_lh)

Warning: Results may not be meaningful for censored models.
[1] "0.51 [0.50;0.51]"
mae_brms(mod_fe_lh)

Warning: Results may not be meaningful for censored models.
[1] "0.40 [0.39;0.40]"
plot_curve(mod_fe_lh, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
lh_days[,paste0("est_free_estradiol_lh", c("", "_se", "_low", "_high"))] <- round(predict(mod_fe_lh, newdata = lh_days, re_formula = NA),2)

Impute P

Backward counting

mod_p_bc = brm(bf(logP | cens(progesterone_cens) ~ s(bc_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle, control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_p_bc')

Warning: Rows containing NAs were excluded from the model.
plot_curve(mod_p_bc, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
bc_days[,paste0("est_progesterone_bc", c("", "_se", "_low", "_high"))] <- round(predict(mod_p_bc, newdata = bc_days, re_formula = NA),2)

Forward counting

mod_p_fc = brm(bf(logP | cens(progesterone_cens)  ~ s(fc_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle, control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_p_fc')

Warning: Rows containing NAs were excluded from the model.
plot_curve(mod_p_fc, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
fc_days[,paste0("est_progesterone_fc", c("", "_se", "_low", "_high"))] <- round(predict(mod_p_fc, newdata = fc_days, re_formula = NA),2)

Relative to LH surge

mod_p_lh = brm(bf(logP | cens(progesterone_cens)  ~ s(lh_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle %>% filter(between(lh_day, -20, 14)), control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_p_lh')

Warning: Rows containing NAs were excluded from the model.
plot_curve(mod_p_lh, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
lh_days[,paste0("est_progesterone_lh", c("", "_se", "_low", "_high"))] <- round(predict(mod_p_lh, newdata = lh_days, re_formula = NA),2)

New PBFW (probability of being in fertile window)

Backward counting

mod_pbfw_bc = brm(bf(fertile_lh ~ s(bc_day) + (1 | id) + (1|id:cycle)),
                data = biocycle, 
                file ='models/mod_pbfw_bc')

Warning: Rows containing NAs were excluded from the model.
plot_curve(mod_pbfw_bc, method = "fitted")

Warning: The following variables in 'conditions' are not part of the model:
'id'
Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
bc_days[,paste0("est_pbfw_bc", c("", "_se", "_low", "_high"))] <- round(predict(mod_pbfw_bc, newdata = bc_days, re_formula = NA),2)

Forward counting

mod_pbfw_fc = brm(bf(fertile_lh ~ s(fc_day)  + (1 | id) + (1|id:cycle)),
                data = biocycle, 
                file ='models/mod_pbfw_fc')

Warning: Rows containing NAs were excluded from the model.
plot_curve(mod_pbfw_fc, method = "fitted")

Warning: The following variables in 'conditions' are not part of the model:
'id'
Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
fc_days[,paste0("est_pbfw_fc", c("", "_se", "_low", "_high"))] <- round(predict(mod_pbfw_fc, newdata = fc_days, re_formula = NA),2)

Relative to LH surge

Not done, because predictor = criterion.

Impute T

Backward counting

mod_t_bc = brm(bf(logT | cens(Testo_cens)  ~ s(bc_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle, control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_t_bc')

Warning: Rows containing NAs were excluded from the model.
loo_R2(mod_t_bc, re_formula = NA)

Warning: Results may not be meaningful for censored models.
Warning: Some Pareto k diagnostic values are too high. See help('pareto-k-diagnostic') for details.
     Estimate  Est.Error       Q2.5      Q97.5
R2 0.05955607 0.01125779 0.03759546 0.08188889
rmse_brms(mod_t_bc)

Warning: Results may not be meaningful for censored models.
[1] "0.17 [0.17;0.18]"
mae_brms(mod_t_bc)

Warning: Results may not be meaningful for censored models.
[1] "0.13 [0.13;0.13]"
plot_curve(mod_t_bc, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
bc_days[,paste0("est_testosterone_bc", c("", "_se", "_low", "_high"))] <- round(predict(mod_t_bc, newdata = bc_days, re_formula = NA),2)

Forward counting

mod_t_fc = brm(bf(logT | cens(Testo_cens)  ~ s(fc_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle, control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_t_fc')

Warning: Rows containing NAs were excluded from the model.
plot_curve(mod_t_fc, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
fc_days[,paste0("est_testosterone_fc", c("", "_se", "_low", "_high"))] <- round(predict(mod_t_fc, newdata = fc_days, re_formula = NA),2)

Relative to LH surge

mod_t_lh = brm(bf(logT | cens(Testo_cens)  ~ s(lh_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle %>% filter(between(lh_day, -20, 14)), control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_t_lh')

Warning: Rows containing NAs were excluded from the model.
plot_curve(mod_t_lh, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
lh_days[,paste0("est_testosterone_lh", c("", "_se", "_low", "_high"))] <- round(predict(mod_t_lh, newdata = lh_days, re_formula = NA),2)

Impute FE

Backward counting

mod_ft_bc = brm(bf(logFT | cens(Testo_cens)  ~ s(bc_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle, control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_ft_bc')

Warning: Rows containing NAs were excluded from the model.
plot_curve(mod_ft_bc, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
bc_days[,paste0("est_free_testosterone_bc", c("", "_se", "_low", "_high"))] <- round(predict(mod_ft_bc, newdata = bc_days, re_formula = NA),2)

Forward counting

mod_ft_fc = brm(bf(logFT | cens(Testo_cens)  ~ s(fc_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle, control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_ft_fc')

Warning: Rows containing NAs were excluded from the model.
plot_curve(mod_ft_fc, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
fc_days[,paste0("est_free_testosterone_fc", c("", "_se", "_low", "_high"))] <- round(predict(mod_ft_fc, newdata = fc_days, re_formula = NA),2)

Relative to LH surge

mod_ft_lh = brm(bf(logFT | cens(Testo_cens)  ~ s(lh_day) + (1 | id) + (1|id:cycle)),  
                data = biocycle %>% filter(between(lh_day, -20, 14)), control = list(adapt_delta = 0.95, max_treedepth = 20),
                file ='models/mod_ft_lh')

Warning: Rows containing NAs were excluded from the model.
plot_curve(mod_ft_lh, method = "fitted")

Warning: Argument 'nsamples' is deprecated. Please use argument
'ndraws' instead.
lh_days[,paste0("est_free_testosterone_lh", c("", "_se", "_low", "_high"))] <- round(predict(mod_ft_lh, newdata = lh_days, re_formula = NA),2)

Label variables

lh_days <- as.data.frame(lh_days)
var_label(fc_days) <- list(fc_day = "Forward-counted cycle day. Counted from last (recalled) menstrual onset, starting at zero.",
                           prc_stirn_fc = "Probability of being in the fertile window determined according to the forward-counted cycle days, probabilities from Gangestad et al. 2016 based on Stirnemann et al. 2013",
                           prc_wcx_fc = "Probability of being in the fertile window determined according to the forward-counted cycle days, probabilities from Gangestad et al. 2016 based on Wilcox et al. 2001",
                           est_estradiol_fc = paste("Imputed serum estradiol level based on the BioCycle study, loo-R =", loo_ci(mod_e_fc)),
                           est_estradiol_fc_se = "Standard error: Imputed serum estradiol.",
                           est_estradiol_fc_low = "Lower 95% CI: Imputed serum estradiol.",
                           est_estradiol_fc_high = "Upper 95% CI: Imputed serum estradiol.",
                           est_free_estradiol_fc = paste("Imputed free serum estradiol level based on the BioCycle study, loo-R =", loo_ci(mod_fe_fc)),
                           est_free_estradiol_fc_se = "Standard error: Imputed free serum estradiol.",
                           est_free_estradiol_fc_low = "Lower 95% CI: Imputed free serum estradiol.",
                           est_free_estradiol_fc_high = "Upper 95% CI: Imputed free serum estradiol.",
                           est_progesterone_fc = paste("Imputed serum progesterone level based on the BioCycle study, loo-R =", loo_ci(mod_p_fc)),
                           est_progesterone_fc_se = "Standard error: Imputed serum progesterone",
                           est_progesterone_fc_low = "Lower 95% CI: Imputed serum progesterone",
                           est_progesterone_fc_high = "Upper 95% CI: Imputed serum progesterone",
                           est_testosterone_fc = paste("Imputed serum testosterone level based on the BioCycle study, loo-R =", loo_ci(mod_t_fc)),
                           est_testosterone_fc_se = "Standard error: Imputed serum testosterone",
                           est_testosterone_fc_low = "Lower 95% CI: Imputed serum testosterone",
                           est_testosterone_fc_high = "Upper 95% CI: Imputed serum testosterone",
                           est_free_testosterone_fc = paste("Imputed free serum testosterone level based on the BioCycle study, loo-R =", loo_ci(mod_ft_fc)),
                           est_free_testosterone_fc_se = "Standard error: Imputed free serum testosterone",
                           est_free_testosterone_fc_low = "Lower 95% CI: Imputed free serum testosterone",
                           est_free_testosterone_fc_high = "Upper 95% CI: Imputed free serum testosterone",
                           est_pbfw_fc = paste("Imputed probability of being in the fertile window based on the BioCycle study (LH test), loo-R =", loo_ci(mod_pbfw_fc)),
                           est_pbfw_fc_se = "Standard error: Imputed probability of being in the fertile window",
                           est_pbfw_fc_low = "Lower 95% CI: Imputed probability of being in the fertile window",
                           est_pbfw_fc_high = "Upper 95% CI: Imputed probability of being in the fertile window"
                           )

Warning: Results may not be meaningful for censored models.

Warning: Results may not be meaningful for censored models.

Warning: Results may not be meaningful for censored models.

Warning: Results may not be meaningful for censored models.
Warning: Some Pareto k diagnostic values are too high. See help('pareto-k-diagnostic') for details.
Warning: Results may not be meaningful for censored models.
Warning: Some Pareto k diagnostic values are too high. See help('pareto-k-diagnostic') for details.
var_label(bc_days) <- list(bc_day = "Backward-counted cycle day. Counted from next observed menstrual onset (0), beginning with -1.",
                           prc_stirn_bc = "Probability of being in the fertile window determined according to the backward-counted cycle days, probabilities from Gangestad et al. 2016 based on Stirnemann et al. 2013",
                           prc_wcx_bc = "Probability of being in the fertile window determined according to the backward-counted cycle days, probabilities from Gangestad et al. 2016 based on Wilcox et al. 2001",
                           est_estradiol_bc = paste("Imputed serum estradiol level based on the BioCycle study, loo-R =", loo_ci(mod_e_bc)),
                           est_estradiol_bc_se = "Standard error: Imputed serum estradiol.",
                           est_estradiol_bc_low = "Lower 95% CI: Imputed serum estradiol.",
                           est_estradiol_bc_high = "Upper 95% CI: Imputed serum estradiol.",
                           est_free_estradiol_bc = paste("Imputed free serum estradiol level based on the BioCycle study, loo-R =", loo_ci(mod_fe_bc)),
                           est_free_estradiol_bc_se = "Standard error: Imputed free serum estradiol.",
                           est_free_estradiol_bc_low = "Lower 95% CI: Imputed free serum estradiol.",
                           est_free_estradiol_bc_high = "Upper 95% CI: Imputed free serum estradiol.",
                           est_progesterone_bc = paste("Imputed serum progesterone level based on the BioCycle study, loo-R =", loo_ci(mod_p_bc)),
                           est_progesterone_bc_se = "Standard error: Imputed serum progesterone",
                           est_progesterone_bc_low = "Lower 95% CI: Imputed serum progesterone",
                           est_progesterone_bc_high = "Upper 95% CI: Imputed serum progesterone",
                           est_testosterone_bc = paste("Imputed serum testosterone level based on the BioCycle study, loo-R =", loo_ci(mod_t_bc)),
                           est_testosterone_bc_se = "Standard error: Imputed serum testosterone",
                           est_testosterone_bc_low = "Lower 95% CI: Imputed serum testosterone",
                           est_testosterone_bc_high = "Upper 95% CI: Imputed serum testosterone",
                           est_free_testosterone_bc = paste("Imputed free serum testosterone level based on the BioCycle study, loo-R =", loo_ci(mod_ft_bc)),
                           est_free_testosterone_bc_se = "Standard error: Imputed free serum testosterone",
                           est_free_testosterone_bc_low = "Lower 95% CI: Imputed free serum testosterone",
                           est_free_testosterone_bc_high = "Upper 95% CI: Imputed free serum testosterone",
                           est_pbfw_bc = paste("Imputed probability of being in the fertile window based on the BioCycle study (LH test), loo-R =", loo_ci(mod_pbfw_bc)),
                           est_pbfw_bc_se = "Standard error: Imputed probability of being in the fertile window",
                           est_pbfw_bc_low = "Lower 95% CI: Imputed probability of being in the fertile window",
                           est_pbfw_bc_high = "Upper 95% CI: Imputed probability of being in the fertile window"
                           )

Warning: Results may not be meaningful for censored models.
Warning: Results may not be meaningful for censored models.

Warning: Results may not be meaningful for censored models.

Warning: Results may not be meaningful for censored models.
Warning: Some Pareto k diagnostic values are too high. See help('pareto-k-diagnostic') for details.
Warning: Results may not be meaningful for censored models.
Warning: Some Pareto k diagnostic values are too high. See help('pareto-k-diagnostic') for details.
var_label(lh_days) <- list(lh_day = "Day relative to LH surge",
                           est_estradiol_lh = paste("Imputed serum estradiol level based on the BioCycle study, loo-R =", loo_ci(mod_e_lh)),
                           est_estradiol_lh_se = "Standard error: Imputed serum estradiol.",
                           est_estradiol_lh_low = "Lower 95% CI: Imputed serum estradiol.",
                           est_estradiol_lh_high = "Upper 95% CI: Imputed serum estradiol.",
                           est_free_estradiol_lh = paste("Imputed free serum estradiol level based on the BioCycle study, loo-R =", loo_ci(mod_fe_lh)),
                           est_free_estradiol_lh_se = "Standard error: Imputed free serum estradiol.",
                           est_free_estradiol_lh_low = "Lower 95% CI: Imputed free serum estradiol.",
                           est_free_estradiol_lh_high = "Upper 95% CI: Imputed free serum estradiol.",
                           est_progesterone_lh = paste("Imputed serum progesterone level based on the BioCycle study, loo-R =", loo_ci(mod_p_lh)),
                           est_progesterone_lh_se = "Standard error: Imputed serum progesterone",
                           est_progesterone_lh_low = "Lower 95% CI: Imputed serum progesterone",
                           est_progesterone_lh_high = "Upper 95% CI: Imputed serum progesterone",
                           est_testosterone_lh = paste("Imputed serum testosterone level based on the BioCycle study, loo-R =", loo_ci(mod_t_lh)),
                           est_testosterone_lh_se = "Standard error: Imputed serum testosterone",
                           est_testosterone_lh_low = "Lower 95% CI: Imputed serum testosterone",
                           est_testosterone_lh_high = "Upper 95% CI: Imputed serum testosterone",
                           est_free_testosterone_lh = paste("Imputed free serum testosterone level based on the BioCycle study, loo-R =", loo_ci(mod_ft_lh)),
                           est_free_testosterone_lh_se = "Standard error: Imputed free serum testosterone",
                           est_free_testosterone_lh_low = "Lower 95% CI: Imputed free serum testosterone",
                           est_free_testosterone_lh_high = "Upper 95% CI: Imputed free serum testosterone",
                           conception_risk_lh = "Risk of conception according to Stern, Kordsmeyer, & Penke 2021 Table 1 weighted averages.",
                           fertile_lh = "Probability of being in the fertile window according to Stern, Kordsmeyer, & Penke 2021 Table 1, taking highest conception risk as the denominator and brought to same mean as Stirnemann numbers."
                           )

Warning: Results may not be meaningful for censored models.
Warning: Results may not be meaningful for censored models.

Warning: Results may not be meaningful for censored models.

Warning: Results may not be meaningful for censored models.
Warning: Some Pareto k diagnostic values are too high. See help('pareto-k-diagnostic') for details.
Warning: Results may not be meaningful for censored models.
Warning: Some Pareto k diagnostic values are too high. See help('pareto-k-diagnostic') for details.

Save merge files

rio::export(fc_days, "merge_files/fc_days.rds")
rio::export(bc_days, "merge_files/bc_days.rds")
rio::export(lh_days, "merge_files/lh_days.rds")
rio::export(fc_days, "merge_files/fc_days.sav")
rio::export(bc_days, "merge_files/bc_days.sav")
rio::export(lh_days, "merge_files/lh_days.sav")
rio::export(fc_days, "merge_files/fc_days.tsv")
rio::export(bc_days, "merge_files/bc_days.tsv")
rio::export(lh_days, "merge_files/lh_days.tsv")

Codebooks

Forward-counted

codebook::compact_codebook(fc_days)

Metadata

Description

Dataset name: results

The dataset has N=40 rows and 27 columns. 40 rows have no missing values on any column.

Metadata for search engines
  • Date published: 2022-03-10
x
fc_day
prc_stirn_fc
prc_wcx_fc
est_estradiol_fc
est_estradiol_fc_se
est_estradiol_fc_low
est_estradiol_fc_high
est_free_estradiol_fc
est_free_estradiol_fc_se
est_free_estradiol_fc_low
est_free_estradiol_fc_high
est_progesterone_fc
est_progesterone_fc_se
est_progesterone_fc_low
est_progesterone_fc_high
est_pbfw_fc
est_pbfw_fc_se
est_pbfw_fc_low
est_pbfw_fc_high
est_testosterone_fc
est_testosterone_fc_se
est_testosterone_fc_low
est_testosterone_fc_high
est_free_testosterone_fc
est_free_testosterone_fc_se
est_free_testosterone_fc_low
est_free_testosterone_fc_high

Codebook table

JSON-LD metadata

The following JSON-LD can be found by search engines, if you share this codebook publicly on the web.

{
  "name": "results",
  "datePublished": "2022-03-10",
  "description": "The dataset has N=40 rows and 27 columns.\n40 rows have no missing values on any column.\n\n\n## Table of variables\nThis table contains variable names, labels, and number of missing values.\nSee the complete codebook for more.\n\n[truncated]\n\n### Note\nThis dataset was automatically described using the [codebook R package](https://rubenarslan.github.io/codebook/) (version 0.9.2).",
  "keywords": ["fc_day", "prc_stirn_fc", "prc_wcx_fc", "est_estradiol_fc", "est_estradiol_fc_se", "est_estradiol_fc_low", "est_estradiol_fc_high", "est_free_estradiol_fc", "est_free_estradiol_fc_se", "est_free_estradiol_fc_low", "est_free_estradiol_fc_high", "est_progesterone_fc", "est_progesterone_fc_se", "est_progesterone_fc_low", "est_progesterone_fc_high", "est_pbfw_fc", "est_pbfw_fc_se", "est_pbfw_fc_low", "est_pbfw_fc_high", "est_testosterone_fc", "est_testosterone_fc_se", "est_testosterone_fc_low", "est_testosterone_fc_high", "est_free_testosterone_fc", "est_free_testosterone_fc_se", "est_free_testosterone_fc_low", "est_free_testosterone_fc_high"],
  "@context": "http://schema.org/",
  "@type": "Dataset",
  "variableMeasured": [
    {
      "name": "fc_day",
      "description": "Forward-counted cycle day. Counted from last (recalled) menstrual onset, starting at zero.",
      "@type": "propertyValue"
    },
    {
      "name": "prc_stirn_fc",
      "description": "Probability of being in the fertile window determined according to the forward-counted cycle days, probabilities from Gangestad et al. 2016 based on Stirnemann et al. 2013",
      "@type": "propertyValue"
    },
    {
      "name": "prc_wcx_fc",
      "description": "Probability of being in the fertile window determined according to the forward-counted cycle days, probabilities from Gangestad et al. 2016 based on Wilcox et al. 2001",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_fc",
      "description": "Imputed serum estradiol level based on the BioCycle study, loo-R = 0.57 [0.55;0.60]",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_fc_se",
      "description": "Standard error: Imputed serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_fc_low",
      "description": "Lower 95% CI: Imputed serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_fc_high",
      "description": "Upper 95% CI: Imputed serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_fc",
      "description": "Imputed free serum estradiol level based on the BioCycle study, loo-R = 0.57 [0.55;0.59]",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_fc_se",
      "description": "Standard error: Imputed free serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_fc_low",
      "description": "Lower 95% CI: Imputed free serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_fc_high",
      "description": "Upper 95% CI: Imputed free serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_fc",
      "description": "Imputed serum progesterone level based on the BioCycle study, loo-R = 0.74 [0.72;0.76]",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_fc_se",
      "description": "Standard error: Imputed serum progesterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_fc_low",
      "description": "Lower 95% CI: Imputed serum progesterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_fc_high",
      "description": "Upper 95% CI: Imputed serum progesterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_pbfw_fc",
      "description": "Imputed probability of being in the fertile window based on the BioCycle study (LH test), loo-R = 0.61 [0.58;0.63]",
      "@type": "propertyValue"
    },
    {
      "name": "est_pbfw_fc_se",
      "description": "Standard error: Imputed probability of being in the fertile window",
      "@type": "propertyValue"
    },
    {
      "name": "est_pbfw_fc_low",
      "description": "Lower 95% CI: Imputed probability of being in the fertile window",
      "@type": "propertyValue"
    },
    {
      "name": "est_pbfw_fc_high",
      "description": "Upper 95% CI: Imputed probability of being in the fertile window",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_fc",
      "description": "Imputed serum testosterone level based on the BioCycle study, loo-R = 0.23 [0.18;0.27]",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_fc_se",
      "description": "Standard error: Imputed serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_fc_low",
      "description": "Lower 95% CI: Imputed serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_fc_high",
      "description": "Upper 95% CI: Imputed serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_fc",
      "description": "Imputed free serum testosterone level based on the BioCycle study, loo-R = 0.18 [0.10;0.23]",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_fc_se",
      "description": "Standard error: Imputed free serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_fc_low",
      "description": "Lower 95% CI: Imputed free serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_fc_high",
      "description": "Upper 95% CI: Imputed free serum testosterone",
      "@type": "propertyValue"
    }
  ]
}`

Backward-counted

codebook::compact_codebook(bc_days)

Metadata

Description

Dataset name: results

The dataset has N=40 rows and 27 columns. 40 rows have no missing values on any column.

Metadata for search engines
  • Date published: 2022-03-10
x
bc_day
prc_stirn_bc
prc_wcx_bc
est_estradiol_bc
est_estradiol_bc_se
est_estradiol_bc_low
est_estradiol_bc_high
est_free_estradiol_bc
est_free_estradiol_bc_se
est_free_estradiol_bc_low
est_free_estradiol_bc_high
est_progesterone_bc
est_progesterone_bc_se
est_progesterone_bc_low
est_progesterone_bc_high
est_pbfw_bc
est_pbfw_bc_se
est_pbfw_bc_low
est_pbfw_bc_high
est_testosterone_bc
est_testosterone_bc_se
est_testosterone_bc_low
est_testosterone_bc_high
est_free_testosterone_bc
est_free_testosterone_bc_se
est_free_testosterone_bc_low
est_free_testosterone_bc_high

Codebook table

JSON-LD metadata

The following JSON-LD can be found by search engines, if you share this codebook publicly on the web.

{
  "name": "results",
  "datePublished": "2022-03-10",
  "description": "The dataset has N=40 rows and 27 columns.\n40 rows have no missing values on any column.\n\n\n## Table of variables\nThis table contains variable names, labels, and number of missing values.\nSee the complete codebook for more.\n\n[truncated]\n\n### Note\nThis dataset was automatically described using the [codebook R package](https://rubenarslan.github.io/codebook/) (version 0.9.2).",
  "keywords": ["bc_day", "prc_stirn_bc", "prc_wcx_bc", "est_estradiol_bc", "est_estradiol_bc_se", "est_estradiol_bc_low", "est_estradiol_bc_high", "est_free_estradiol_bc", "est_free_estradiol_bc_se", "est_free_estradiol_bc_low", "est_free_estradiol_bc_high", "est_progesterone_bc", "est_progesterone_bc_se", "est_progesterone_bc_low", "est_progesterone_bc_high", "est_pbfw_bc", "est_pbfw_bc_se", "est_pbfw_bc_low", "est_pbfw_bc_high", "est_testosterone_bc", "est_testosterone_bc_se", "est_testosterone_bc_low", "est_testosterone_bc_high", "est_free_testosterone_bc", "est_free_testosterone_bc_se", "est_free_testosterone_bc_low", "est_free_testosterone_bc_high"],
  "@context": "http://schema.org/",
  "@type": "Dataset",
  "variableMeasured": [
    {
      "name": "bc_day",
      "description": "Backward-counted cycle day. Counted from next observed menstrual onset (0), beginning with -1.",
      "@type": "propertyValue"
    },
    {
      "name": "prc_stirn_bc",
      "description": "Probability of being in the fertile window determined according to the backward-counted cycle days, probabilities from Gangestad et al. 2016 based on Stirnemann et al. 2013",
      "@type": "propertyValue"
    },
    {
      "name": "prc_wcx_bc",
      "description": "Probability of being in the fertile window determined according to the backward-counted cycle days, probabilities from Gangestad et al. 2016 based on Wilcox et al. 2001",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_bc",
      "description": "Imputed serum estradiol level based on the BioCycle study, loo-R = 0.68 [0.66;0.69]",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_bc_se",
      "description": "Standard error: Imputed serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_bc_low",
      "description": "Lower 95% CI: Imputed serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_bc_high",
      "description": "Upper 95% CI: Imputed serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_bc",
      "description": "Imputed free serum estradiol level based on the BioCycle study, loo-R = 0.68 [0.66;0.69]",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_bc_se",
      "description": "Standard error: Imputed free serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_bc_low",
      "description": "Lower 95% CI: Imputed free serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_bc_high",
      "description": "Upper 95% CI: Imputed free serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_bc",
      "description": "Imputed serum progesterone level based on the BioCycle study, loo-R = 0.83 [0.81;0.84]",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_bc_se",
      "description": "Standard error: Imputed serum progesterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_bc_low",
      "description": "Lower 95% CI: Imputed serum progesterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_bc_high",
      "description": "Upper 95% CI: Imputed serum progesterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_pbfw_bc",
      "description": "Imputed probability of being in the fertile window based on the BioCycle study (LH test), loo-R = 0.69 [0.66;0.71]",
      "@type": "propertyValue"
    },
    {
      "name": "est_pbfw_bc_se",
      "description": "Standard error: Imputed probability of being in the fertile window",
      "@type": "propertyValue"
    },
    {
      "name": "est_pbfw_bc_low",
      "description": "Lower 95% CI: Imputed probability of being in the fertile window",
      "@type": "propertyValue"
    },
    {
      "name": "est_pbfw_bc_high",
      "description": "Upper 95% CI: Imputed probability of being in the fertile window",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_bc",
      "description": "Imputed serum testosterone level based on the BioCycle study, loo-R = 0.24 [0.19;0.28]",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_bc_se",
      "description": "Standard error: Imputed serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_bc_low",
      "description": "Lower 95% CI: Imputed serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_bc_high",
      "description": "Upper 95% CI: Imputed serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_bc",
      "description": "Imputed free serum testosterone level based on the BioCycle study, loo-R = 0.22 [0.17;0.27]",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_bc_se",
      "description": "Standard error: Imputed free serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_bc_low",
      "description": "Lower 95% CI: Imputed free serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_bc_high",
      "description": "Upper 95% CI: Imputed free serum testosterone",
      "@type": "propertyValue"
    }
  ]
}`

Relative to LH surge

codebook::compact_codebook(lh_days)

Metadata

Description

Dataset name: results

The dataset has N=31 rows and 23 columns. 31 rows have no missing values on any column.

Metadata for search engines
  • Date published: 2022-03-10
x
conception_risk_lh
lh_day
fertile_lh
est_estradiol_lh
est_estradiol_lh_se
est_estradiol_lh_low
est_estradiol_lh_high
est_free_estradiol_lh
est_free_estradiol_lh_se
est_free_estradiol_lh_low
est_free_estradiol_lh_high
est_progesterone_lh
est_progesterone_lh_se
est_progesterone_lh_low
est_progesterone_lh_high
est_testosterone_lh
est_testosterone_lh_se
est_testosterone_lh_low
est_testosterone_lh_high
est_free_testosterone_lh
est_free_testosterone_lh_se
est_free_testosterone_lh_low
est_free_testosterone_lh_high

Codebook table

JSON-LD metadata

The following JSON-LD can be found by search engines, if you share this codebook publicly on the web.

{
  "name": "results",
  "datePublished": "2022-03-10",
  "description": "The dataset has N=31 rows and 23 columns.\n31 rows have no missing values on any column.\n\n\n## Table of variables\nThis table contains variable names, labels, and number of missing values.\nSee the complete codebook for more.\n\n[truncated]\n\n### Note\nThis dataset was automatically described using the [codebook R package](https://rubenarslan.github.io/codebook/) (version 0.9.2).",
  "keywords": ["conception_risk_lh", "lh_day", "fertile_lh", "est_estradiol_lh", "est_estradiol_lh_se", "est_estradiol_lh_low", "est_estradiol_lh_high", "est_free_estradiol_lh", "est_free_estradiol_lh_se", "est_free_estradiol_lh_low", "est_free_estradiol_lh_high", "est_progesterone_lh", "est_progesterone_lh_se", "est_progesterone_lh_low", "est_progesterone_lh_high", "est_testosterone_lh", "est_testosterone_lh_se", "est_testosterone_lh_low", "est_testosterone_lh_high", "est_free_testosterone_lh", "est_free_testosterone_lh_se", "est_free_testosterone_lh_low", "est_free_testosterone_lh_high"],
  "@context": "http://schema.org/",
  "@type": "Dataset",
  "variableMeasured": [
    {
      "name": "conception_risk_lh",
      "description": "Risk of conception according to Stern, Kordsmeyer, & Penke 2021 Table 1 weighted averages.",
      "@type": "propertyValue"
    },
    {
      "name": "lh_day",
      "description": "Day relative to LH surge",
      "@type": "propertyValue"
    },
    {
      "name": "fertile_lh",
      "description": "Probability of being in the fertile window according to Stern, Kordsmeyer, & Penke 2021 Table 1, taking highest conception risk as the denominator and brought to same mean as Stirnemann numbers.",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_lh",
      "description": "Imputed serum estradiol level based on the BioCycle study, loo-R = 0.72 [0.70;0.74]",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_lh_se",
      "description": "Standard error: Imputed serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_lh_low",
      "description": "Lower 95% CI: Imputed serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_estradiol_lh_high",
      "description": "Upper 95% CI: Imputed serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_lh",
      "description": "Imputed free serum estradiol level based on the BioCycle study, loo-R = 0.72 [0.70;0.74]",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_lh_se",
      "description": "Standard error: Imputed free serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_lh_low",
      "description": "Lower 95% CI: Imputed free serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_estradiol_lh_high",
      "description": "Upper 95% CI: Imputed free serum estradiol.",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_lh",
      "description": "Imputed serum progesterone level based on the BioCycle study, loo-R = 0.87 [0.85;0.88]",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_lh_se",
      "description": "Standard error: Imputed serum progesterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_lh_low",
      "description": "Lower 95% CI: Imputed serum progesterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_progesterone_lh_high",
      "description": "Upper 95% CI: Imputed serum progesterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_lh",
      "description": "Imputed serum testosterone level based on the BioCycle study, loo-R = 0.27 [0.22;0.31]",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_lh_se",
      "description": "Standard error: Imputed serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_lh_low",
      "description": "Lower 95% CI: Imputed serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_testosterone_lh_high",
      "description": "Upper 95% CI: Imputed serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_lh",
      "description": "Imputed free serum testosterone level based on the BioCycle study, loo-R = 0.27 [0.21;0.31]",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_lh_se",
      "description": "Standard error: Imputed free serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_lh_low",
      "description": "Lower 95% CI: Imputed free serum testosterone",
      "@type": "propertyValue"
    },
    {
      "name": "est_free_testosterone_lh_high",
      "description": "Upper 95% CI: Imputed free serum testosterone",
      "@type": "propertyValue"
    }
  ]
}`