laser.generic.SEIRS

`laser.generic.SEIRS`

Export required components for an SEIRS model.

Agents transition from Susceptible to Exposed upon infection, with an incubation duration. Agents transition from Exposed to Infectious after the incubation period and are infectious for a duration. Agents transition from Infectious to Recovered after the infectious period. Agents transition from Recovered back to Susceptible upon waning immunity after the waning duration

`laser.generic.SEIRS.Exposed(model, expdurdist, infdurdist, expdurmin=1, infdurmin=1, validating=False)`

Exposed Component for SEIR/SEIRS Models with Explicit Incubation Period

This component handles the incubation phase in models where agents must transition from an 'exposed' (E) state to 'infectious' (I) after a delay. It supports custom incubation and infectious duration distributions and handles both initialization and per-tick dynamics.

Agents transition from Exposed to Infectious when their incubation timer (etimer) expires. Tracks number of agents becoming infectious each tick in model.nodes.newly_infectious.

Responsibilities: - Initializes exposed individuals at time 0 (if provided in the scenario) - Assigns and tracks per-agent incubation timers (etimer) - Transitions agents from EXPOSED to INFECTIOUS when etimer == 0 - Assigns new infection timers (itimer) upon becoming infectious - Updates patch-level EXPOSED (E) and INFECTIOUS case counts - Provides validation hooks for state and timer consistency

Required Inputs: - model.scenario.E: initial count of exposed individuals per node (optional) - expdurdist: callable returning sampled incubation durations - infdurdist: callable returning sampled infectious durations - expdurmin: minimum incubation period (default 1 day) - infdurmin: minimum infectious period (default 1 day)

Outputs: - model.people.etimer: agent-level incubation timer - model.nodes.E[t, i]: number of exposed individuals at time t in node i - model.nodes.newly_infectious[t, i]: number of newly infectious cases per node per day

Validation: - Ensures consistency between individual states and etimer values - Ensures that agents becoming infectious have valid itimer values assigned - Prevents agents with expired etimer from remaining in EXPOSED state

Step Behavior

For each agent: - Decrease etimer - If etimer == 0, change state to INFECTIOUS and assign itimer - Update model.nodes.E and model.nodes.I counts accordingly

Plotting: The plot() method provides a time series of exposed individuals per node and total across all nodes.

Example

model.components = [ SIR.Susceptible(model), Exposed(model, expdurdist, infdurdist), SIR.Infectious(model, infdurdist), ... ]

`laser.generic.SEIRS.Infectious(model, infdurdist, wandurdist, infdurmin=1, wandurmin=1, validating=False)`

Infectious Component for SIRS/SEIRS Models (Recovery with Waning Immunity)

This component manages infectious individuals in models where recovery confers temporary immunity, after which agents become susceptible again (SIRS/SEIRS).

Agents transition from Infectious to Recovered after the infectious period (itimer). Set the waning immunity timer (rtimer) upon recovery. Tracks number of agents recovering each tick in model.nodes.newly_recovered.

Responsibilities: - Initializes infectious agents from model.scenario.I - Assigns and tracks infectious timers (itimer) per agent - Transitions agents from INFECTIOUS to RECOVERED when itimer == 0 - Assigns a waning immunity timer (rtimer) upon recovery - Updates patch-level state: • I[t, i]: current infectious count • R[t, i]: current recovered count • recovered[t, i]: number of agents recovering on tick t

Required Inputs: - model.scenario.I: number of initially infected agents per node - infdurdist: function that samples the infectious duration distribution - wandurdist: function that samples the waning immunity duration distribution - infdurmin: minimum infectious period (default = 1 day) - wandurmin: minimum duration of immunity (default = 1 day)

Outputs: - model.people.itimer: days remaining in the infectious state - model.people.rtimer: days remaining in the recovered state - model.nodes.I, model.nodes.R: counts per node per tick - model.nodes.recovered[t]: number of recoveries recorded on tick t

Step Behavior: - Infectious agents decrement their itimer - When itimer == 0, agents become recovered and receive an rtimer - Patch-level totals are updated - Downstream components (e.g., Recovered) handle rtimer countdown and eventual return to SUSCEPTIBLE

Validation: - Ensures timer consistency and population accounting - Confirms correct infectious-to-recovered transitions - Can be chained with recovery and waning components for full SIRS/SEIRS loops

Plotting: Two plots are provided: 1. Infected counts per node 2. Total infected and recovered counts across time

Example

model.components = [ SIR.Susceptible(model), InfectiousIRS(model, infdurdist, wandurdist), Exposed(model, ...), Recovered(model), ]

`laser.generic.SEIRS.Infectious.step(tick)`

Step function for the Infected component.

Parameters:

Name	Type	Description	Default
`tick`	`int`	The current tick of the simulation.	required

`laser.generic.SEIRS.Recovered(model, wandurdist, wandurmin=1, validating=False)`

Recovered Component for SIRS/SEIRS Models (Waning Immunity)

This component manages agents in the recovered state in models where immunity is temporary. It supports per-agent recovery timers, enabling individuals to return to the susceptible state after a configurable waning period. This is essential for SEIRS/SIRS model dynamics.

Agents transition from Recovered back to Susceptible after the waning immunity period (rtimer). Tracks number of agents losing immunity each tick in model.nodes.newly_waned.

Responsibilities: - Initializes agents in the RECOVERED state using model.scenario.R - Assigns rtimer values to track the duration of immunity - Decrements rtimer each tick; transitions agents to SUSCEPTIBLE when rtimer == 0 - Updates patch-level counts: • R[t, i]: number of recovered individuals in node i at time t • waned[t, i]: number of agents who re-entered susceptibility on time step t

Required Inputs: - model.scenario.R: initial number of recovered individuals per node - wandurdist: a function sampling the waning immunity duration distribution - wandurmin: minimum duration of immunity (default = 1 time step)

Outputs: - model.people.rtimer: per-agent countdown to immunity expiration - model.nodes.R: recovered count per patch per timestep - model.nodes.waned: number of immunity losses per patch per tick

Step Behavior: - Agents with state == RECOVERED decrement rtimer - When rtimer == 0, they return to SUSCEPTIBLE - R and S counts are updated to reflect this transition - waned[t] logs the number of agents who lost immunity on time step t

Validation: - Ensures population conservation and consistency between agent states and patch totals - Detects unexpected changes in R or invalid transitions

Plotting: The plot() method provides two views: 1. Per-node recovered trajectories 2. Total recovered and waned agents over time

Example

model.components = [ SIR.Susceptible(model), SEIRS.Infectious(model, infdurdist, wandurdist), Exposed(model, ...), RecoveredRS(model, wandurdist), ]

`laser.generic.SEIRS.Susceptible(model, validating=False)`

Susceptible Component for Patch-Based Agent-Based Models (S, SI, SIS, SIR, SEIR, etc.)

This component initializes and tracks the count of susceptible individuals (S) in a spatially structured agent-based model. It is compatible with all standard LASER disease progression models that include a "susceptible" state.

Responsibilities: - Initializes agent-level properties: • nodeid: Patch ID of each agent (uint16) • state: Infection state (int8), defaulting to State.SUSCEPTIBLE - Initializes node-level property: • S[t, i]: Susceptible count in node i at time t - At each timestep, propagates the susceptible count forward (S[t+1] = S[t]), unless modified by other components (e.g., exposure, births). - Validates consistency between patch-level susceptible counts and agent-level state.

Usage: Add this component early in the component list for any model with SUSCEPTIBLE agents, typically before transmission or exposure components. Compatible with: - SIR.Transmission - SIR.Exposure - SIR.Infectious - SIR.Recovered - Custom SEIRS extensions

Requires: - model.people: A LaserFrame for all agents - model.nodes: Patch-level state - model.scenario: Input DataFrame with population and optionally S columns - model.params.nticks: Number of simulation ticks

Validation: - Ensures consistency of susceptible counts before and after each step - Prevents unintentional state drift by validating against agent state values

Output: - model.nodes.S: A (nticks+1, num_nodes) array of susceptible counts - Optional plotting via plot() for visual inspection of per-node and total S

Step Behavior

For tick t: S[t+1] = S[t] # Unless explicitly modified by other components

This component does not alter agent states directly but serves as a synchronized counter and validator of susceptible individuals.

Example

model.components = [ SIR.Susceptible(model), SIR.Transmission(model, ...), SIR.Exposure(model), SIR.Infectious(model, ...), SIR.Recovered(model), ]

`laser.generic.SEIRS.Transmission(model, expdurdist, expdurmin=1, seasonality=None, validating=False)`

Transmission component for an SEIR/SEIRS model with S -> E transition and incubation duration.

This component simulates the transition from `SUSCEPTIBLE` to `EXPOSED` in models
where infection includes an incubation period before agents become infectious.
It handles stochastic exposure based on per-node force of infection (FOI), and
assigns individual incubation timers to newly exposed agents.

Agents transition from Susceptible to Exposed based on force of infection.
Sets newly exposed agents' infection timers (etimer) based on `expdurdist` and `expdurmin`.
Tracks number of new infections each tick in `model.nodes.newly_infected`.

Responsibilities:
- Computes force of infection `λ = β * (I / N)` at each tick per node
- Adjusts FOI using `model.network` for inter-node transmission coupling. Required but can be nullified by filling with all zeros.
- Applies FOI to susceptible agents to determine exposure
- Assigns incubation durations (`etimer`) to each newly exposed agent
- Updates node-level counts for `S` and `E` and logs daily incidence

Required Inputs:
- `model.params.beta`: global transmission rate
- `model.network`: [n x n] matrix for FOI migration
- `expdurdist(tick, node)`: callable that samples the exposure/incubation duration distribution
- `expdurmin`: minimum incubation period (default = 1)

Outputs:
- `model.nodes.forces[t, i]`: computed FOI in node `i` at tick `t`
- `model.nodes.incidence[t, i]`: new exposures per node per day
- `model.people.etimer`: per-agent incubation countdown

Step Behavior:
- Computes FOI (`λ`) for each node
- Optionally applies inter-node infection pressure via `model.network`
- Converts FOI into a Bernoulli probability using: `p = 1 - exp(-λ)`
- Infects susceptible agents probabilistically
- Updates state and records incidence

Validation:
- Validates consistency between agent states and patch-level counts before and after tick
- Confirms that `incidence[t] == E[t+1] - E[t]`

Plotting:
The `plot()` method shows per-node FOI (`λ`) trajectories over time.

Example:
    model.components = [
        SIR.Susceptible(model),
        TransmissionSE(model, expdurdist),
        Exposed(model, ...),
        InfectiousIR(model, ...),
        Recovered(model),
    ]

Initializes the TransmissionSE component.

Parameters:

Name	Type	Description	Default
`model`	`Model`	The epidemiological model instance.	required
`expdurdist`	`Callable[[int, int], float]`	A function that returns the incubation duration for a given tick and node.	required
`expdurmin`	`int`	Minimum incubation duration.	`1`
`seasonality`	`Union[ValuesMap, ndarray]`	Seasonality modifier for transmission rate. Defaults to None.	`None`
`validating`	`bool`	Enable component-level validation. Defaults to False.	`False`