Skip to content

GitLab

Commit f379a4e2 authored by St. Elmo's avatar St. Elmo
Browse files

fixed formatting

parent 08b0b781
Hide whitespace changes
Inline Side-by-side
docs/make.jl
View file @ f379a4e2
......@@ -3,28 +3,29 @@ using Documenter, COBREXA
# download core model
download("http://bigg.ucsd.edu/static/models/e_coli_core.json", "e_coli_core.json")
makedocs(modules = [COBREXA],
clean = false,
sitename = "COBREXA.jl",
format = Documenter.HTML(
# Use clean URLs, unless built as a "local" build
prettyurls = !("local" in ARGS),
assets = ["assets/favicon.ico"],
highlights = ["yaml"],
),
authors = "The developers of COBREXA.jl",
linkcheck = !("skiplinks" in ARGS),
pages = [
"Home" => "index.md",
"Functions" => "functions.md",
"How to contribute" => "howToContribute.md",
"Model Structure" => "model_structure.md",
"Model IO" => "io.md",
"Model Construction" => "model_construction.md",
"Optimization Based Analysis Tools" => "basic_analysis.md",
"Sampling Tools" => "sampling_tools.md",
],
)
makedocs(
modules = [COBREXA],
clean = false,
sitename = "COBREXA.jl",
format = Documenter.HTML(
# Use clean URLs, unless built as a "local" build
prettyurls = !("local" in ARGS),
assets = ["assets/favicon.ico"],
highlights = ["yaml"],
),
authors = "The developers of COBREXA.jl",
linkcheck = !("skiplinks" in ARGS),
pages = [
"Home" => "index.md",
"Functions" => "functions.md",
"How to contribute" => "howToContribute.md",
"Model Structure" => "model_structure.md",
"Model IO" => "io.md",
"Model Construction" => "model_construction.md",
"Optimization Based Analysis Tools" => "basic_analysis.md",
"Sampling Tools" => "sampling_tools.md",
],
)
# delete core model
rm("e_coli_core.json")
\ No newline at end of file
rm("e_coli_core.json")
src/COBREXA.jl
View file @ f379a4e2
......@@ -27,7 +27,8 @@ include(joinpath("header", "types.jl"))
const PKG_ROOT_DIR = normpath(joinpath(@__DIR__, ".."))
include_dependency(joinpath(PKG_ROOT_DIR, "Project.toml"))
const COBREXA_VERSION = VersionNumber(Pkg.TOML.parsefile(joinpath(PKG_ROOT_DIR, "Project.toml"))["version"])
const COBREXA_VERSION =
VersionNumber(Pkg.TOML.parsefile(joinpath(PKG_ROOT_DIR, "Project.toml"))["version"])
c = Base.text_colors
tx = c[:normal] # text
......@@ -58,17 +59,34 @@ loadSource(["io", "reconstruction", "analysis", "sampling"], @__DIR__)
# export functions
∅ = Metabolite("∅") # for exchange reactions
export speye, LinearModel, nReactions, nMetabolites, nCouplingConstraints,
addReaction, addReactions, removeReactions, changeBounds!,
addCouplingConstraints!, addCouplingConstraints,
removeCouplingConstraints!, removeCouplingConstraints,
changeCouplingBounds!, changeCouplingBounds,
verifyConsistency, findExchangeReactions, findExchangeMetabolites,
solveLP, loadModel, loadSBMLModel, writeModel,
fluxBalanceAnalysis, fluxVariabilityAnalysis, parFVA,
convertToExportable,
# from CobraTools
export speye,
LinearModel,
nReactions,
nMetabolites,
nCouplingConstraints,
addReaction,
addReactions,
removeReactions,
changeBounds!,
addCouplingConstraints!,
addCouplingConstraints,
removeCouplingConstraints!,
removeCouplingConstraints,
changeCouplingBounds!,
changeCouplingBounds,
verifyConsistency,
findExchangeReactions,
findExchangeMetabolites,
solveLP,
loadModel,
loadSBMLModel,
writeModel,
fluxBalanceAnalysis,
fluxVariabilityAnalysis,
parFVA,
convertToExportable,
# from CobraTools
∅,
# base
CobraModel,
......@@ -113,4 +131,4 @@ export speye, LinearModel, nReactions, nMetabolites, nCouplingConstraints,
test_samples,
achr
end # module
\ No newline at end of file
end # module
src/analysis/fba.jl
View file @ f379a4e2
......@@ -3,10 +3,15 @@ Convert LinearModel to the JuMP model, place objectives and the equality
constraint.
"""
function makeOptimizationModel(model::LinearModel, optimizer)
m,n = size(model.S)
m, n = size(model.S)
optimization_model = JuMP.Model(optimizer)
@variable(optimization_model, x[i=1:n], lower_bound=model.xl[i], upper_bound=model.xu[i])
@variable(
optimization_model,
x[i = 1:n],
lower_bound = model.xl[i],
upper_bound = model.xu[i]
)
@objective(optimization_model, Max, model.c' * x)
@constraint(optimization_model, model.S * x .== model.b)
return (optimization_model, x)
......@@ -22,10 +27,10 @@ s.t. S x = b
"""
function fluxBalanceAnalysis(model::LinearModel, optimizer)
optimization_model, x = makeOptimizationModel(model, optimizer)
JuMP.optimize!(optimization_model)
return (optimization_model, x)
# TODO we might like this to take optimization model and return x, and let
# the user convert LinearModel to JuMP model. That would make the API much
# more orthogonal.
optimization_model, x = makeOptimizationModel(model, optimizer)
JuMP.optimize!(optimization_model)
return (optimization_model, x)
# TODO we might like this to take optimization model and return x, and let
# the user convert LinearModel to JuMP model. That would make the API much
# more orthogonal.
end
src/analysis/fva.jl
View file @ f379a4e2
......@@ -12,32 +12,32 @@ where Z₀:= cᵀx₀ is the objective value of an optimal solution to the assoc
FBA problem
"""
function fluxVariabilityAnalysis(model::LinearModel, optimizer)
(m, n) = size(model.S)
return fluxVariabilityAnalysis(model, collect(1:n), optimizer)
(m, n) = size(model.S)
return fluxVariabilityAnalysis(model, collect(1:n), optimizer)
end
function fluxVariabilityAnalysis(model::LinearModel, reactions::Vector{Int64}, optimizer)
(maximum(reactions) > length(model.rxns)) && error("Index exceeds number of reactions.")
γ = 1.
fluxes = zeros(length(reactions), 2)
(maximum(reactions) > length(model.rxns)) && error("Index exceeds number of reactions.")
γ = 1.0
fluxes = zeros(length(reactions), 2)
(optimization_model, x₀) = fluxBalanceAnalysis(model::LinearModel, optimizer)
Z₀ = JuMP.objective_value(optimization_model)
x = all_variables(optimization_model)
@constraint(optimization_model, model.c' * x ≥ γ * Z₀)
(optimization_model, x₀) = fluxBalanceAnalysis(model::LinearModel, optimizer)
Z₀ = JuMP.objective_value(optimization_model)
x = all_variables(optimization_model)
@constraint(optimization_model, model.c' * x ≥ γ * Z₀)
for i in eachindex(reactions)
sense = MOI.MIN_SENSE
@objective(optimization_model, sense, x[reactions[i]])
JuMP.optimize!(optimization_model)
fluxes[i, 1] = JuMP.objective_value(optimization_model)
for i in eachindex(reactions)
sense = MOI.MIN_SENSE
@objective(optimization_model, sense, x[reactions[i]])
JuMP.optimize!(optimization_model)
fluxes[i, 1] = JuMP.objective_value(optimization_model)
sense = MOI.MAX_SENSE
JuMP.set_objective_sense(optimization_model, sense)
JuMP.optimize!(optimization_model)
fluxes[i, 2] = JuMP.objective_value(optimization_model)
end
return fluxes
sense = MOI.MAX_SENSE
JuMP.set_objective_sense(optimization_model, sense)
JuMP.optimize!(optimization_model)
fluxes[i, 2] = JuMP.objective_value(optimization_model)
end
return fluxes
end
function parFVA_add_constraint(model, c, x, Z0, gamma)
......@@ -58,22 +58,32 @@ function parFVA(model::LinearModel, reactions::Vector{Int}, optimizer, workers)
throw(ArgumentError("reactions contain an out-of-bounds index"))
end
gamma = 1. #TODO parametrize?
gamma = 1.0 #TODO parametrize?
(optimization_model, x0) = fluxBalanceAnalysis(model::LinearModel, optimizer)
Z0 = JuMP.objective_value(optimization_model)
# make a JuMP optimization model
map(fetch, save_at.(workers, :cobrexa_parfva_model, Ref(:(begin
optmodel, x = COBREXA.makeOptimizationModel($model, $optimizer)
COBREXA.parFVA_add_constraint(optmodel, $(model.c), x, $Z0, $gamma)
optmodel
end))))
map(
fetch,
save_at.(
workers,
:cobrexa_parfva_model,
Ref(:(
begin
optmodel, x = COBREXA.makeOptimizationModel($model, $optimizer)
COBREXA.parFVA_add_constraint(optmodel, $(model.c), x, $Z0, $gamma)
optmodel
end
)),
),
)
# schedule FVA parts parallely using pmap
fluxes = dpmap(
rid -> :(COBREXA.parFVA_get_opt(cobrexa_parfva_model, $rid)),
CachingPool(workers),
[-reactions reactions])
[-reactions reactions],
)
# free the data on workers
map(fetch, remove_from.(workers, :cobrexa_parfva_model))
......
src/base/solver.jl
View file @ f379a4e2
......@@ -2,14 +2,19 @@
Use JuMP to solve an instance of LinearModel
"""
function solveLP(model::LinearModel, optimizer)
m, n = size(model.S)
m, n = size(model.S)
optimization_model = JuMP.Model(optimizer)
@variable(optimization_model, x[i=1:n], lower_bound=model.xl[i], upper_bound=model.xu[i])
@objective(optimization_model, Min, model.c' * x)
@constraint(optimization_model, model.S * x .== model.b)
optimization_model = JuMP.Model(optimizer)
@variable(
optimization_model,
x[i = 1:n],
lower_bound = model.xl[i],
upper_bound = model.xu[i]
)
@objective(optimization_model, Min, model.c' * x)
@constraint(optimization_model, model.S * x .== model.b)
JuMP.optimize!(optimization_model)
JuMP.optimize!(optimization_model)
return (optimization_model, x)
return (optimization_model, x)
end
src/base/utilities.jl
View file @ f379a4e2
......@@ -4,14 +4,29 @@
speye(n) = sparse(1.0I, n, n)
function Base.isequal(model1::LinearModel, model2::LinearModel)
return model1.S == model2.S && model1.b == model2.b && model1.C == model2.C &&
model1.cl == model2.cl && model1.cu == model2.cu &&
return model1.S == model2.S &&
model1.b == model2.b &&
model1.C == model2.C &&
model1.cl == model2.cl &&
model1.cu == model2.cu &&
model1.c == model2.c &&
model1.xl == model2.xl && model1.xu == model2.xu &&
model1.rxns == model2.rxns && model1.mets == model2.mets
model1.xl == model2.xl &&
model1.xu == model2.xu &&
model1.rxns == model2.rxns &&
model1.mets == model2.mets
end
function Base.copy(model::LinearModel)
return LinearModel(model.S, model.b, model.C, model.cl, model.cu,
model.c, model.xl, model.xu, model.rxns, model.mets)
return LinearModel(
model.S,
model.b,
model.C,
model.cl,
model.cu,
model.c,
model.xl,
model.xu,
model.rxns,
model.mets,
)
end
src/header/header.jl
View file @ f379a4e2
......@@ -4,7 +4,7 @@ loadSource(srcDirs, root=@__DIR__)
Load source files
"""
function loadSource(srcDirs, rootDir=@__DIR__, testMode=false)
function loadSource(srcDirs, rootDir = @__DIR__, testMode = false)
returnFlag = []
for d in srcDirs
srcDir = joinpath(rootDir, d)
......@@ -24,4 +24,4 @@ function loadSource(srcDirs, rootDir=@__DIR__, testMode=false)
end
end
return returnFlag
end
\ No newline at end of file
end
src/header/types.jl
View file @ f379a4e2
const VT = Union{Array{Float64,1}, SparseVector{Float64,Int64}}
const MT = Union{AbstractMatrix, SparseMatrixCSC{Float64,Int64}}
const ST = Union{Array{String,1}, SparseVector{String,Int64}}
const VT = Union{Array{Float64,1},SparseVector{Float64,Int64}}
const MT = Union{AbstractMatrix,SparseMatrixCSC{Float64,Int64}}
const ST = Union{Array{String,1},SparseVector{String,Int64}}
"""
A linear optimization problem of the form:
......@@ -11,26 +11,27 @@ s.t. S x = b
xₗ ≤ x ≤ xᵤ
```
"""
mutable struct LinearModel{M<:MT, V<:VT, K<:ST}
S ::M
b ::V
C ::M
cl ::V
cu ::V
c ::V
xl ::V
xu ::V
rxns ::K
mets ::K
mutable struct LinearModel{M<:MT,V<:VT,K<:ST}
S::M
b::V
C::M
cl::V
cu::V
c::V
xl::V
xu::V
rxns::K
mets::K
function LinearModel(
S ::M,
b ::V,
c ::V,
xl ::V,
xu ::V,
rxns ::K,
mets ::K) where {V<:VT,M<:MT,K<:ST}
S::M,
b::V,
c::V,
xl::V,
xu::V,
rxns::K,
mets::K,
) where {V<:VT,M<:MT,K<:ST}
sS = sparse(S)
......@@ -46,16 +47,17 @@ mutable struct LinearModel{M<:MT, V<:VT, K<:ST}
end
function LinearModel(
S ::M1,
b ::V,
C ::M2,
cl ::V,
cu ::V,
c ::V,
xl ::V,
xu ::V,
rxns ::K,
mets ::K) where {V<:VT,M1<:MT,M2<:MT,K<:ST}
S::M1,
b::V,
C::M2,
cl::V,
cu::V,
c::V,
xl::V,
xu::V,
rxns::K,
mets::K,
) where {V<:VT,M1<:MT,M2<:MT,K<:ST}
checkInputDimensions(S, b, C, cl, cu, c, xl, xu, rxns, mets)
......@@ -68,7 +70,18 @@ mutable struct LinearModel{M<:MT, V<:VT, K<:ST}
sxl = sparse(xl)
sxu = sparse(xu)
new{SparseMatrixCSC{Float64,Int64}, SparseVector{Float64,Int64}, Array{String,1}}(sS, sb, sC, scl, scu, sc, sxl, sxu, rxns, mets)
new{SparseMatrixCSC{Float64,Int64},SparseVector{Float64,Int64},Array{String,1}}(
sS,
sb,
sC,
scl,
scu,
sc,
sxl,
sxu,
rxns,
mets,
)
end
end
......
src/io/reader.jl
View file @ f379a4e2
......@@ -21,7 +21,7 @@ Convert a dictionary read from a MAT file to LinearModel
function convertToLinearModel(model::Dict)
modelKeys = ["S", "b", "c", "ub", "lb"]
for key = modelKeys
for key in modelKeys
if !(key in keys(model))
error("No variable $key found in the MAT file.")
end
......
src/io/sbml.jl
View file @ f379a4e2
......@@ -18,14 +18,18 @@ function loadSBMLModel(filename::String)
# error if not.
unit = lbu[1][2]
getvalue = (val,_)::Tuple -> val
getvalue = (val, _)::Tuple -> val
getunit = (_, unit)::Tuple -> unit
allunits = unique([getunit.(lbu) getunit.(ubu)])
if length(allunits) != 1
throw(DomainError(allunits,
"The SBML file uses multiple units; loading needs conversion"))
throw(
DomainError(
allunits,
"The SBML file uses multiple units; loading needs conversion",
),
)
end
return LinearModel(S, b, c, getvalue.(lbu), getvalue.(ubu), rxns, mets)
......
src/io/writer.jl
View file @ f379a4e2
......@@ -5,7 +5,7 @@ NB: Does NOT export general inequality constraints (eg coupling)
See also: `MAT.jl`
"""
function writeModel(filePath::String, model::LinearModel, varName::String="model")
function writeModel(filePath::String, model::LinearModel, varName::String = "model")
matwrite(filePath, Dict(varName => convertToExportable(model)))
end
......@@ -14,11 +14,13 @@ Convert a LinearModel to exportable format
SparseVectors are not written and read properly, SparseMatrix is okay
"""
function convertToExportable(model::LinearModel)
return Dict("S" => model.S,
"b" => Array(model.b),
"c" => Array(model.c),
"ub" => Array(model.xu),
"lb" => Array(model.xl),
"rxns" => model.rxns,
"mets" => model.mets)
return Dict(
"S" => model.S,
"b" => Array(model.b),
"c" => Array(model.c),
"ub" => Array(model.xu),
"lb" => Array(model.xl),
"rxns" => model.rxns,
"mets" => model.mets,
)
end
src/reconstruction/coupling.jl
View file @ f379a4e2
"""
Verifies that vectors and matrices have the expected dimensions.
"""
function checkCouplingConstraintsInputDimensions(m::LinearModel,
C::M,
cl::V,
cu::V) where {M<:MT,V<:VT}
length(cu) == length(cl) || throw(DimensionMismatch("`cl` and `cu` don't have the same size"))
size(C) == (length(cl), nReactions(m)) || throw(DimensionMismatch("`C` size should be length(cl) x the number of reactions in m"))
function checkCouplingConstraintsInputDimensions(
m::LinearModel,
C::M,
cl::V,
cu::V,
) where {M<:MT,V<:VT}
length(cu) == length(cl) ||
throw(DimensionMismatch("`cl` and `cu` don't have the same size"))
size(C) == (length(cl), nReactions(m)) || throw(
DimensionMismatch("`C` size should be length(cl) x the number of reactions in m"),
)
end
"""
......@@ -16,36 +21,44 @@ Add constraints of the following form to a LinearModel:
cₗ ≤ C x ≤ cᵤ
```
"""
function addCouplingConstraints(m::LinearModel,
c::V,
cl::AbstractFloat,
cu::AbstractFloat) where {V<:VT}
return addCouplingConstraints(m, sparse(reshape(c, (1, length(c)))), sparse([cl]), sparse([cu]))
function addCouplingConstraints(
m::LinearModel,
c::V,
cl::AbstractFloat,
cu::AbstractFloat,
) where {V<:VT}
return addCouplingConstraints(
m,
sparse(reshape(c, (1, length(c)))),
sparse([cl]),
sparse([cu]),
)
end
function addCouplingConstraints(m::LinearModel,
C::M,
cl::V,
cu::V) where {M<:MT,V<:VT}
function addCouplingConstraints(m::LinearModel, C::M, cl::V, cu::V) where {M<:MT,V<:VT}
newLp = deepcopy(m)
addCouplingConstraints!(newLp, C, cl, cu)
return newLp
end
function addCouplingConstraints!(m::LinearModel,
c::V,
cl::AbstractFloat,
cu::AbstractFloat) where {V<:VT}
return addCouplingConstraints!(m, sparse(reshape(c, (1, length(c)))), sparse([cl]), sparse([cu]))
function addCouplingConstraints!(
m::LinearModel,
c::V,
cl::AbstractFloat,
cu::AbstractFloat,
) where {V<:VT}
return addCouplingConstraints!(
m,
sparse(reshape(c, (1, length(c)))),
sparse([cl]),
sparse([cu]),
)
end
function addCouplingConstraints!(m::LinearModel,
C::M,
cl::V,
cu::V) where {M<:MT,V<:VT}
function addCouplingConstraints!(m::LinearModel, C::M, cl::V, cu::V) where {M<:MT,V<:VT}
C = sparse(C)
cl = sparse(cl)
......@@ -62,29 +75,25 @@ end
"""
Removes a set of coupling constraints from a LinearModel.
"""
function removeCouplingConstraints(m::LinearModel,
constraint::Int64)
function removeCouplingConstraints(m::LinearModel, constraint::Int64)
return removeCouplingConstraints(m, [constraint])
end
function removeCouplingConstraints(m::LinearModel,
constraints::Array{Int64, 1})
function removeCouplingConstraints(m::LinearModel, constraints::Array{Int64,1})
newModel = deepcopy(m)
removeCouplingConstraints!(newModel, constraints)
return newModel
end
function removeCouplingConstraints!(m::LinearModel,
constraint::Int64)
function removeCouplingConstraints!(m::LinearModel, constraint::Int64)
removeCouplingConstraints!(m, [constraint])
end
function removeCouplingConstraints!(m::LinearModel,
constraints::Array{Int64, 1})
toBeKept = filter(e->e ∉ constraints, 1:nCouplingConstraints(m))
function removeCouplingConstraints!(m::LinearModel, constraints::Array{Int64,1})
toBeKept = filter(e -> e ∉ constraints, 1:nCouplingConstraints(m))
m.C = m.C[toBeKept, :]
m.cl = m.cl[toBeKept]
m.cu = m.cu[toBeKept]
......@@ -102,20 +111,26 @@ end