API Documentation

run_model(f, ::Type{Model{M}}, input::AbstractInput) where M
run_model(f, ::Type{Model{M}}, input::AbstractInput, state::AbstractState) where M

Run a model and send Frames to callback f. The type parameter M should be a model, being a module with both initial_state! and step! defined.

The second version with explicit state is used by H5Writer so we can perform an additional action between creating the initial state and starting the model run (saving metadata to the HDF5 file).

run_model(::Type{Model{M}}, input::AbstractInput, filename::AbstractString) where M

Run a model and write output to HDF5.

run_model(Model{ALCAP}, ALCAP.Example.INPUT, "example.h5")

run_model(::Type{Model{M}}, input::AbstractInput, output::AbstractOutput) where M

Run a model and save the output to output.

Box{BoundaryType}(grid_size, phys_scale)

Stores the spatial properties of the simulation grid, where grid_size is a tuple of two integers and phys_scale is a Quantity of dimension Length.

box_axes(box::Box)

Return the x and y components of the box grid in meters. The grid starts at 0.0u"m" and runs to (box.grid_size .- 1) .* box.phys_scale.

TimeProperties(t0, Δt, steps, write_interval)

Stores properties of the time integration. Here, t0 and Δt should be a Quantity, steps is the number of integration steps, and write_interval the number of steps between writing output.

time_axis(time::TimeProperties)

Retrieve the time values for which output was/will be written. Returns a range.

n_steps(input)

Returns the number of steps from a given input.

step_ca(box, facies)

Creates a propagator for the state, updating the celullar automaton in place.

Contract: the state should have ca::Matrix{Int} and ca_priority::Vector{Int} members.

disintegrator(input) -> f!

Prepares the disintegration step. Returns a function f!(state::State). The returned function modifies the state, popping sediment from the sediment_buffer and returns an array of Amount.

transporter(input::Input) -> f

Prepares the transportation step. Returns a function f(state::State, active_layer), transporting the active layer, returning a transported Amount of sediment.

add_data_set(out::T, name::Symbol, spec::OutputSpec)

Add a data set to the output object.

frame_writer(input::AbstractInput, out::T)

Returns a function (idx::Int, state::AbstractState).

Write the state of the simulation to the output object. It is the responsibility of the implementation to choose to write or not based on the set write interval.

The default implementation writes the state for all output data sets, and calls write_sediment_thickness.

new_output(::Type{T}, input)

Create a new output object of type T, given input.

state_writer(input::AbstractInput, out::T)

Returns a function (idx::Int, state::AbstractState).

Write the state of the simulation to the output object. It is the responsibility of the implementation to choose to write or not based on the set write interval.

The default implementation writes the state for all output data sets, and calls write_sediment_thickness.

write_deposition(out::T, name::Symbol, idx::Int, data::AbstractArray{Amount, dim}) where {T, dim}

See write_sediment_thickness. Should accept 1, 2, and 3 dimensional arrays, corresponding to writing column, slice or volume data. (first axis is facies, then x and y)

write_disintegration(out::T, name::Symbol, idx::Int, data::AbstractArray{Amount, dim}) where {T, dim}

See write_sediment_thickness. Should accept 1, 2, and 3 dimensional arrays, corresponding to writing column, slice or volume data. (first axis is facies, then x and y)

write_production(out::T, name::Symbol, idx::Int, data::AbstractArray{Amount, dim}) where {T, dim}

See write_sediment_thickness. Should accept 1, 2, and 3 dimensional arrays, corresponding to writing column, slice or volume data. (first axis is facies, then x and y)

write_sediment_thickness(out::T, name::Symbol, idx::Int, data::AbstractArray{Amount, dim}) where {T, dim}

Write the sediment thickness to the output object. The idx should be corrected for write interval. That is, idx should range from 1 to n_writes for the named data set. This function should be implemented for 0, 1, and 2 dimensional arrays, corresponding to writing column, slice or volume data.

If your output object type doesn't conform to the standard CarboKitten data layout, you may choose to not implement this function and implement state_writer and frame_writer instead. The same goes for write_production, write_disintegration and write_deposition.

set_attribute(out::T, name::String, value::Any)

Set an attribute in the output object.

run_model(::Type{Model{M}}, input::AbstractInput, filename::AbstractString) where M

Run a model and write output to HDF5.

run_model(Model{ALCAP}, ALCAP.Example.INPUT, "example.h5")

enumerate_seq(s)

Enumerates an iterator of sequences, such that the following equivalence holds:

enumerate(flatten(s)) == flatten(enumerate_seq(s))

find_ranges(v::AbstractVector{Bool})

Take a vector of bools, returns an iterator over all ranges for which the vector is true.

skeleton(bitmap::AbstractMatrix{Bool})

Computes the skeleton of a bitmap, i.e. reduces features with some thickness to a set of line segments. This function is designed with stratigraphic application in mind: we scan each row in the bitmap for connected regions, then link neighbouring regions when they overlap. The result is a graph that represents hiatus in the sediment accumulation.

Returns a tuple of vertices and edges, where vertices is a vector of 2-tuples and edges is a nx2 matrix of indices into the vertices.

stencil!(f, Boundary, Size, out, inp...)

Performs stencil operation. The function f should take a number of abstract arrays (same as the number of inp values), and return a single value of the same type as elements in out. The Size parameter should be a type parameter size of the stencil, e.g. Size(3, 3) to get a 3 by 3 stencil.

Prefer to use this version over the older implementations.

age_depth_model(sediment_accumulation_curve::Vector)
age_depth_model(sediment_accumulation_curve::DataFrame)

Compute the ADM from the SAC. Implemented as:

reverse ∘ accumulate(min) ∘ reverse

The DataFrame version selects SAC columns, transformed into ADM.

data_export(spec::CSV, header::Header, data)

Exports data to CSV. Here, data should be a collection or iterable of DataColumn.

extract_sac(header::Header, data::DataColumn)

Extract Sediment Accumumlation Curve (SAC) from the data. The SAC is directly copied from data.sediment_thickness. Returns a DataFrame with time and sac_<n> columns where <n> is in the range 1:length(grid_locations).

extract_sc(header::Header, data::DataColumn)

Extract Stratigraphic Column (SC) from the data. Returns a DataFrame with time and sc<n> columns where <n> is in the range 1:length(grid_locations).

extract_wd(header::Header, data::DataColumn)

Extract the water depth from the data. Returns a DataFrame with time and wd<n> columns where <n> is in the range 1:length(grid_locations).

stratigraphic_column(header::Header, column::DataColumn, facies::Int)

Compute the Stratigraphic Column for a given grid position loc and facies index. Returns an Array{Quantity, 2} where the Quantity is in units of meters.

unitful_headers(df::DataFrame)

Gets a string representation for all column names including their unit. Returns a Vector{String}.

write_unitful_csv(io::IO, df::DataFrame)

Write a CSV from a DataFrame with Unitful units. The units will be represented in the CSV header, and stripped from the individual values.

ustrip(df::DataFrame)

Strip units from a DataFrame. Returns a new DataFrame.

transport!(box, diffusivity, wave_velocity,
           C, w, dC)

Computes dC given a box, diffusivity constant in units of m/Myr, wave_velocity is a function of water depth, returning both velocity in units of m/Myr, and shear in units of 1/Myr, which should be the derivative of the velocity w.r.t. water depth. C is the concentration of entrained sediment, w the water depth, and dC the output derivative of C.

transport(box, diffusivity, wave_velocity, wave_shear,
           C, w)

Non-mutating version of transport!. Allocates and returns dC.

denudation(box, param, state)

Computes the denudation for a single time-step, given denudation parameters param and a simulation state state. param should have a DenudationType type and state should contain the height property and sealevel.

Returns denudation mass in units of meters.

denudation(box::Box, param::DenudationType, water_depth, slope, facies)

Computes the amount of denudation. This function is called on a pixel by pixel basis, so all arguments can be assumed to be scalar. The param argument should be of a subtype of DenudationType containing all the input parameters for this specific denudation model.

redistribution()

Takes state, water_depth in meters and denudation_mass as a 3D array (facies, x and y coordinates) in units of meters.

module NoDenudation

Doesn't do any denudation: used for testing purposes.