"Nested Partially Latent Class Models"

Documentation for package ‘baker’ version 1.0.2

Help Pages

-- A --

add_meas_BrS_case_Nest_Slice	add likelihood for a BrS measurement slice among cases (conditional dependence)
add_meas_BrS_case_Nest_Slice_jags	add likelihood for a BrS measurement slice among cases (conditional dependence)
add_meas_BrS_case_NoNest_reg_discrete_predictor_Slice_jags	add likelihood component for a BrS measurement slice among cases
add_meas_BrS_case_NoNest_reg_Slice_jags	add likelihood component for a BrS measurement slice among cases
add_meas_BrS_case_NoNest_Slice	add a likelihood component for a BrS measurement slice among cases (conditional independence)
add_meas_BrS_case_NoNest_Slice_jags	add a likelihood component for a BrS measurement slice among cases (conditional independence)
add_meas_BrS_ctrl_Nest_Slice	add likelihood for a BrS measurement slice among controls (conditional independence)
add_meas_BrS_ctrl_NoNest_reg_discrete_predictor_Slice_jags	add a likelihood component for a BrS measurement slice among controls
add_meas_BrS_ctrl_NoNest_reg_Slice_jags	add a likelihood component for a BrS measurement slice among controls
add_meas_BrS_ctrl_NoNest_Slice	add a likelihood component for a BrS measurement slice among controls (conditional independence)
add_meas_BrS_param_Nest_reg_Slice_jags	add parameters for a BrS measurement slice among cases and controls
add_meas_BrS_param_Nest_Slice	add parameters for a BrS measurement slice among cases and controls (conditional dependence)
add_meas_BrS_param_Nest_Slice_jags	add parameters for a BrS measurement slice among cases and controls (conditional dependence)
add_meas_BrS_param_NoNest_reg_discrete_predictor_Slice_jags	add parameters for a BrS measurement slice among cases and controls
add_meas_BrS_param_NoNest_reg_Slice_jags	add parameters for a BrS measurement slice among cases and controls
add_meas_BrS_param_NoNest_Slice	add parameters for a BrS measurement slice among cases and controls (conditional independence)
add_meas_BrS_param_NoNest_Slice_jags	add parameters for a BrS measurement slice among cases and controls (conditional independence)
add_meas_BrS_subclass_Nest_Slice	add subclass indicators for a BrS measurement slice among cases and controls (conditional independence)
add_meas_SS_case	add likelihood for a SS measurement slice among cases (conditional independence)
add_meas_SS_param	add parameters for a SS measurement slice among cases (conditional independence)
as.matrix_or_vec	convert one column data frame to a vector
assign_model	Interpret the specified model structure

-- B --

baker	baker: Bayesian Analytic Kit for Etiology Research
beta_parms_from_quantiles	Pick parameters in the Beta distribution to match the specified range
beta_plot	Plot beta density
bin2dec	Convert a 0/1 binary-coded sequence into decimal digits

-- C --

check_dir_create	check existence and create folder if non-existent
clean_combine_subsites	Combine subsites in raw PERCH data set
clean_perch_data	Clean PERCH data
combine_data_nplcm	combine multiple data_nplcm (useful when simulating data from regression models)
compute_logOR_single_cause	Calculate marginal log odds ratios
compute_marg_PR_nested_reg	compute positive rates for nested model with subclass mixing weights that are the same across 'Jcause' classes for each person (people may have different weights.)
compute_marg_PR_nested_reg_array	compute positive rates for nested model with subclass mixing weights that are the same across 'Jcause' classes for each person (people may have different weights.)
create_bugs_regressor_Eti	create regressor summation equation used in regression for etiology
create_bugs_regressor_FPR	create regressor summation equation used in regression for FPR

-- D --

data_nplcm_noreg	Simulated dataset that is structured in the format necessary for an 'nplcm()' without regression
data_nplcm_reg_nest	Simulated dataset that is structured in the format necessary for an 'nplcm()' with regression
delete_start_with	Deletes a pattern from the start of a string, or each of a vector of strings.
dm_Rdate_Eti	Make etiology design matrix for dates with R format.
dm_Rdate_FPR	Make FPR design matrix for dates with R format.

-- E --

expit	expit function
extract_data_raw	Import Raw PERCH Data 'extract_data_raw' imports and converts the raw data to analyzable format

-- G --

get_coverage	Obtain coverage status from a result folder
get_direct_bias	Obtain direct bias that measure the discrepancy of a posterior distribution of pie and a true pie.
get_fitted_mean_nested	get fitted mean for nested model with subclass mixing weights that are the same among cases
get_fitted_mean_no_nested	get model fitted mean for conditional independence model
get_individual_data	get individual data
get_individual_prediction	get individual prediction (Bayesian posterior)
get_latent_seq	get index of latent status
get_marginal_rates_nested	get marginal TPR and FPR for nested model
get_marginal_rates_no_nested	get marginal TPR and FPR for no nested model
get_metric	Obtain Integrated Squared Aitchison Distance, Squared Bias and Variance (both on Central Log-Ratio transformed scale) that measure the discrepancy of a posterior distribution of pie and a true pie.
get_pEti_samp	get etiology samples by names (no regression)
get_plot_num	get the plotting positions (numeric) for the fitted means; 3 positions for each cell
get_plot_pos	get a list of measurement index where to look for data
get_postsd	Obtain posterior standard deviation from a result folder
get_top_pattern	get top patterns from a slice of bronze-standard measurement

-- H --

H	Shannon entropy for multivariate discrete data
has_non_basis	test if a formula has terms not created by [s_date_Eti() or 's_date_FPR()'

-- I --

I2symb	Convert 0/1 coding to pathogen/combinations
Imat2cat	Convert a matrix of binary indicators to categorical variables
init_latent_jags_multipleSS	Initialize individual latent status (for 'JAGS')
insert_bugfile_chunk_noreg_etiology	insert distribution for latent status code chunk into .bug file
insert_bugfile_chunk_noreg_meas	Insert measurement likelihood (without regression) code chunks into .bug model file
insert_bugfile_chunk_reg_discrete_predictor_etiology	insert etiology regression for latent status code chunk into .bug file; discrete predictors
insert_bugfile_chunk_reg_discrete_predictor_nonest_meas	Insert measurement likelihood (with regression; discrete) code chunks into .bug model file
insert_bugfile_chunk_reg_etiology	insert etiology regression for latent status code chunk into .bug file
insert_bugfile_chunk_reg_nest_meas	Insert measurement likelihood (nested model+regression) code chunks into .bug model file
insert_bugfile_chunk_reg_nonest_meas	Insert measurement likelihood (with regression) code chunks into .bug model file
is.error	Test for 'try-error' class
is_discrete	Check if covariates are discrete
is_intercept_only	check if the formula is intercept only
is_jags_folder	See if a result folder is obtained by JAGS
is_length_all_one	check if a list has elements all of length one

-- J --

jags2_baker

Run 'JAGS' from R

-- L --

line2user	convert line to user coordinates
loadOneName	load an object from .RDATA file
logit	logit function
logOR	calculate pairwise log odds ratios
logsumexp	log sum exp trick
lookup_quality	Get position to store in data_nplcm$Mobs:

-- M --

make_filename	Create new file name
make_foldername	Create new folder name
make_list	Takes any number of R objects as arguments and returns a list whose names are derived from the names of the R objects.
make_meas_object	Make measurement slice
make_numbered_list	Make a list with numbered names
make_template	make a mapping template for model fitting
marg_H	Shannon entropy for binary data
match_cause	Match latent causes that might have the same combo but different specifications
merge_lists	For a list of many sublists each of which has matrices as its member, we combine across the many sublists to produce a final list
my_reorder	Reorder the measurement dimensions to match the order for display

-- N --

NA2dot	convert 'NA' to '.'
nplcm	Fit nested partially-latent class models (highest-level wrapper function)
nplcm_fit_NoReg	Fit nested partially-latent class model (low-level)
nplcm_fit_Reg_discrete_predictor_NoNest	Fit nested partially-latent class model with regression (low-level)
nplcm_fit_Reg_Nest	Fit nested partially-latent class model with regression (low-level)
nplcm_fit_Reg_NoNest	Fit nested partially-latent class model with regression (low-level)
nplcm_read_folder	Read data and other model information from a folder that stores model results.
null_as_zero	Convert 'NULL' to zero.

-- O --

order_post_eti	order latent status by posterior mean
overall_uniform	specify overall uniform (symmetric Dirichlet distribution) for etiology prior

-- P --

parse_nplcm_reg	parse regression components (either false positive rate or etiology regression) for fitting npLCM; Only use this when formula is not 'NULL'.
pathogen_category_perch	pathogens and their categories in PERCH study (virus or bacteria)
pathogen_category_simulation	Hypothetical pathogens and their categories (virus or bacteria)
plot.nplcm	'plot.nplcm' plot the results from 'nplcm()'.
plot_BrS_panel	Plot bronze-standard (BrS) panel
plot_case_study	visualize the PERCH etiology regression with a continuous covariate
plot_check_common_pattern	Posterior predictive checking for the nested partially class models - frequent patterns in the BrS data. (for multiple folders)
plot_check_pairwise_SLORD	Posterior predictive checking for nested partially latent class models - pairwise log odds ratio (only for bronze-standard data)
plot_etiology_regression	visualize the etiology regression with a continuous covariate
plot_etiology_strat	visualize the etiology estimates for each discrete levels
plot_leftmost	plotting the labels on the left margin for panels plot
plot_logORmat	Visualize pairwise log odds ratios (LOR) for data that are available in both cases and controls
plot_panels	Plot three-panel figures for nested partially-latent model results
plot_pie_panel	Plot etiology (pie) panel
plot_SS_panel	Plot silver-standard (SS) panel
plot_subwt_regression	visualize the subclass weight regression with a continuous covariate
print.nplcm	'print.nplcm' summarizes the results from 'nplcm()'.
print.summary.nplcm.no_reg	Compact printing of 'nplcm()' model fits
print.summary.nplcm.reg_nest	Compact printing of 'nplcm()' model fits
print.summary.nplcm.reg_nest_strat	Compact printing of 'nplcm()' model fits
print.summary.nplcm.reg_nonest	Compact printing of 'nplcm()' model fits
print.summary.nplcm.reg_nonest_strat	Compact printing of 'nplcm()' model fits

-- R --

read_meas_object	Read measurement slices
rvbern	Sample a vector of Bernoulli variables.

-- S --

set_prior_tpr_BrS_NoNest	Set true positive rate (TPR) prior ranges for bronze-standard (BrS) data
set_prior_tpr_SS	Set true positive rate (TPR) prior ranges for silver-standard data.
set_strat	Stratification setup by covariates
show_dep	Show function dependencies
show_individual	get an individual's data from the output of 'clean_perch_data()'
simulate_brs	Simulate Bronze-Standard (BrS) Data
simulate_latent	Simulate Latent Status:
simulate_nplcm	Simulate data from nested partially-latent class model (npLCM) family
simulate_ss	Simulate Silver-Standard (SS) Data
softmax	softmax
subset_data_nplcm_by_index	subset data from the output of 'clean_perch_data()'
summarize_BrS	summarize bronze-standard data
summarize_SS	silver-standard data summary
summary.nplcm	'summary.nplcm' summarizes the results from 'nplcm()'.
symb2I	Convert names of pathogen/combinations into 0/1 coding
sym_diff_month	get symmetric difference of months from two vector of R-format dates
s_date_Eti	Make Etiology design matrix for dates with R format.
s_date_FPR	Make false positive rate (FPR) design matrix for dates with R format.

-- T --

tsb	generate stick-breaking prior (truncated) from a vector of random probabilities

-- U --

unfactor	Convert factor to numeric without losing information on the label
unique_cause	get unique causes, regardless of the actual order in combo
unique_month	Get unique month from Date

-- V --

visualize_case_control_matrix	Visualize matrix for a quantity measured on cases and controls (a single number)
visualize_season	visualize trend of pathogen observation rate for NPPCR data (both cases and controls)

-- W --

write.model	function to write bugs model (copied from R2WinBUGS)
write_model_NoReg	Write .bug model file for model without regression
write_model_Reg_discrete_predictor_NoNest	Write .bug model file for regression model without nested subclasses
write_model_Reg_Nest	Write '.bug' model file for regression model WITH nested subclasses
write_model_Reg_NoNest	Write .bug model file for regression model without nested subclasses