Functions for modelbuilding \n
build_model
: takes the riverdata, handles the other functions and
invokes stan_lm
build_model(riverdata, variables = ask_for_variables(riverdata), with_interaction = TRUE) ask_for_variables(riverdata) process_model_riverdata(riverdata, variables) create_formula(variables, with_interaction = FALSE)
riverdata | a list with riverdata (hygiene + physical data) |
---|---|
variables | character. Selected variables for the model |
with_interaction | logical. Formula with interactions? Default set to TRUE |
Returns a model of the riverdata.
Returns a character-vector with the chosen model variables
Returns a data.frame with data for hygiene and chosen variables
Returns parsed model-formula. (Like model$formula)
Build the model from hygiene data and physical data like flow, rain,
wwtp. Asks for user input to select variables. Computes the data.frame with
data for hygiene and chosen variables and creates a formula of the form:
Q*(K + R)
while multiple Qs will be multiplied, multiple Ks and Rs
will be added.
ask_for_variables
: Internal function. Quite time consuming
process_model_riverdata
: Internal usage
create_formula
: Internal usage
if (FALSE) variables <- c("e.coli","q_havel",...) lm(formula = eval(create_formula(variables)), data = process_model_riverdata(riverdata, variables))#> Error in create_formula(variables): object 'variables' not found#> log_e.coli ~ q_havel + ka_ruhleben + r_berlin #> <environment: 0x12999210>#> e.coli ~ r_mitte + r_charlottenburg + r_spandau #> <environment: 0x129fac68>