Used to calculate observation details based on
cohorts created with system_define_cohort
system_od_general(pest, cfg, estimation = TRUE, details = FALSE)If estimation is TRUE then the output is a matrix of observation details of the format:
od$pred = [TIME, OBS, PRED, VAR, OUTPUT, COHORT] The values are the observed (OBS) data, predicted
values (PRED) and variance (VAR) at the given TIME. The columns OUTPUT and
COHORT can be used for sorting. These should be unique numbers.
When estimation is FALSE we output od$pred is a data frame with the
following headings:
od$pred = [TIME, OBS, PRED, VAR, SMOOTH, OUTPUT, COHORT] The TIME, OBS, PRED and VAR are the same as those listed above. The SMOOTH
variable is FALSE for rows that correspond to records in the dataset and
TRUE when the PRED represents the smooth predictions. The OUTPUT and COHORT
columns here are text values used when defining the cohorts.
Also the od$all list item is created with all of the simulation information
stored for each cohort:
od$all = [ts.time, ts.ts1, ... ts.tsn, pred, name, cohort]
tstime - timescale of the system
ts.ts1, ... ts.tsn - timescales defined in the system
pred - smooth prediction
name - state or output name corresponding to the prediction
cohort - name of the cohort for these predictions
Lastly the field isgood will be set to FALSE if any problems are encountered, and TRUE if everything worked.
od$isgood = TRUEvector of parameters to be estimated
ubiquity system object
TRUE when called during an estimation and FALSE when called to test objective function or generate observation information for plotting
TRUE to display information about cohorts as they are simulated (useful for debugging when passed through system_simulate_estimation_results)
system_define_cohort and system_simulate_estimation_results