expected_obs_from.stan File Reference

Go to the source code of this file.

Functions
vector	expected_obs_from_index (int i, array[] vector imp_obs, array[,] int rdlurd, vector srdlh, matrix refp_lh, array[] int dpmfs, int ref_p, int rep_h, int ref_as_p, int g, int t, int l, vector refnp_lh, int ref_np, int p, int rep_agg_p, array[,,] int rep_agg_n_selected, array[,,,] int rep_agg_selected_idx)
vector	expected_obs_from_snaps (int start, int end, array[] vector imp_obs, array[,] int rdlurd, vector srdlh, matrix refp_lh, array[] int dpmfs, int ref_p, int rep_h, int ref_as_p, array[] int sl, array[] int csl, array[] int sg, array[] int st, int n, vector refnp_lh, int ref_np, array[] int sdmax, array[] int csdmax, int rep_agg_p, array[,,] int rep_agg_n_selected, array[,,,] int rep_agg_selected_idx)

Function Documentation

expected_obs_from_index()

vector expected_obs_from_index	(	int	i,
		array[]vector	imp_obs,
		array int	rdlurd[,],
		vector	srdlh,
		matrix	refp_lh,
		array[]int	dpmfs,
		int	ref_p,
		int	rep_h,
		int	ref_as_p,
		int	g,
		int	t,
		int	l,
		vector	refnp_lh,
		int	ref_np,
		int	p,
		int	rep_agg_p,
		array int	rep_agg_n_selected[,,],
		array int	rep_agg_selected_idx[,,,] )

Calculate expected observations for a given index

Computes the expected observations for a specific time index by combining final observed/imputed expected observation with time effects. This function integrates various hazard effects and reference adjustments to produce the expected observations.

Parameters

start	Start index for the range of interest.
end	End index for the range of interest.
obs	Array of observed data, stored as a flat vector.
sl	Array of start index for each observation period.
csl	Array of cumulative start indexes.
nsl	Array of next start indexes, indicating the start of the next period.
cnsl	Array of cumulative next start indexes.
obs_lookup	Array of indexes linking observed data to modeled expectations.
imp_obs	Array of imputed observed data, organized by group and time.
sg	Array indicating group index for each observation.
st	Array indicating time index for each observation.
rdlurd	Array indicating reporting dates lookup reference data.
srdlh	Vector of standardized reporting date log hazards.
refp_lh	Matrix of reference date logit hazards.
dpmfs	Array of indices for accessing reference date effects in refp_lh.
ref_p	Binary flag for reference date effects presence.
rep_h	Binary flag for reporting hazard effects presence.
ref_as_p	Binary flag indicating if reference date input should be treated as probability.
phi	Array of dispersion parameters for negative binomial distribution.
model_obs	Binary flag indicating if a negative binomial model is used.
refnp_lh	Vector of non-parametric reference log hazards.
ref_np	Binary flag for non-parametric reference effects presence.
sdmax	Array of maximum start dates for each period.
csdmax	Array of cumulative start dates.
rep_agg_p	Binary flag indicating whether reporting probabilities should be aggregated for structural reporting patterns (e.g., fixed reporting cycles).
rep_agg_n_selected	3D array of counts of selected indices per row for probability aggregation, with dimensions [groups, times, max_delay].
rep_agg_selected_idx	4D array of column indices for aggregation, with dimensions [groups, times, max_delay, max_delay].
i	Index for accessing elements in various arrays.
g	Group index for data access.
t	Time index for data access.
l	Length parameter for data processing.
p	Position parameter for data access.

Returns

A vector of expected observations for the given index.

Note

This function performs several steps:

1. Retrieves the final observed/imputed expected observation for the given group and time.
2. Aggregates various hazard effects using combine_logit_hazards.
3. Combines the final expected observation with the time-varying effects to compute the expected observations using expected_obs.

Dependencies:

• combine_logit_hazards
• expected_obs

Definition at line 32 of file expected_obs_from.stan.

expected_obs_from_snaps()

vector expected_obs_from_snaps	(	int	start,
		int	end,
		array[]vector	imp_obs,
		array int	rdlurd[,],
		vector	srdlh,
		matrix	refp_lh,
		array[]int	dpmfs,
		int	ref_p,
		int	rep_h,
		int	ref_as_p,
		array[]int	sl,
		array[]int	csl,
		array[]int	sg,
		array[]int	st,
		int	n,
		vector	refnp_lh,
		int	ref_np,
		array[]int	sdmax,
		array[]int	csdmax,
		int	rep_agg_p,
		array int	rep_agg_n_selected[,,],
		array int	rep_agg_selected_idx[,,,] )

Calculate expected observations for a set of indexes

Computes expected observations for a range of time indexes, effectively aggregating individual calculations over a specified range. This function considers various factors like hazard effects, reference adjustments, and grouping to produce expected observations for each index within the range.

Parameters

start	Start index for the range of interest.
end	End index for the range of interest.
obs	Array of observed data, stored as a flat vector.
sl	Array of start index for each observation period.
csl	Array of cumulative start indexes.
nsl	Array of next start indexes, indicating the start of the next period.
cnsl	Array of cumulative next start indexes.
obs_lookup	Array of indexes linking observed data to modeled expectations.
imp_obs	Array of imputed observed data, organized by group and time.
sg	Array indicating group index for each observation.
st	Array indicating time index for each observation.
rdlurd	Array indicating reporting dates lookup reference data.
srdlh	Vector of standardized reporting date log hazards.
refp_lh	Matrix of reference date logit hazards.
dpmfs	Array of indices for accessing reference date effects in refp_lh.
ref_p	Binary flag for reference date effects presence.
rep_h	Binary flag for reporting hazard effects presence.
ref_as_p	Binary flag indicating if reference date input should be treated as probability.
phi	Array of dispersion parameters for negative binomial distribution.
model_obs	Binary flag indicating if a negative binomial model is used.
refnp_lh	Vector of non-parametric reference log hazards.
ref_np	Binary flag for non-parametric reference effects presence.
sdmax	Array of maximum start dates for each period.
csdmax	Array of cumulative start dates.
rep_agg_p	Binary flag indicating whether reporting probabilities should be aggregated for structural reporting patterns (e.g., fixed reporting cycles).
rep_agg_n_selected	3D array of counts of selected indices per row for probability aggregation, with dimensions [groups, times, max_delay].
rep_agg_selected_idx	4D array of column indices for aggregation, with dimensions [groups, times, max_delay, max_delay].
n	Number of discrete points for observations.

Returns

A vector of expected observations across the specified range of indexes.

Note

This function executes the following steps:

1. Iteratively processes each index in the specified range.
2. For each index, it identifies the relevant group, time, and length parameters necessary for the calculation.
3. Calls expected_obs_from_index for each index to compute the expected observations.
4. Aggregates these individual expected observations into a single vector corresponding to the entire range.

Dependencies:

• expected_obs_from_index

Definition at line 94 of file expected_obs_from.stan.