Merge remote-tracking branch 'origin/master'

docs/src/analysis/bifurcation.md

@@ -41,7 +41,7 @@ p = [-60,120,0.0,2,4,20,-1.2,18]
 Then we use the following command to build the PyDSTool ODE:
 
 ```julia
-dsargs = build_ode(f,u0,tspan)
+dsargs = build_ode(f,u0,tspan,p)
 ```
 
 Now we need to build the continuation type. Following the setup of PyDSTool's

docs/src/analysis/parameter_estimation.md

@@ -394,6 +394,14 @@ p = [1.5,1.0,3.0,1.0]
 prob = ODEProblem(f2,u0,tspan,p)
 ```
 
+We can build an objective function and solve the multiple parameter version just as before:
+
+```julia
+cost_function = build_loss_objective(prob,Tsit5(),CostVData(t,data),
+                                      maxiters=10000,verbose=false)
+result_bfgs = Optim.optimize(cost_function, [1.3,0.8,2.8,1.2], Optim.BFGS())
+```
+
 To solve it using LeastSquaresOptim.jl, we use the `build_lsoptim_objective` function:
 
 ```julia

docs/src/analysis/sensitivity.md

@@ -81,10 +81,10 @@ for the Lotka-Volterra equations by:
 ```julia
 f = @ode_def_nohes LotkaVolterraSensitivity begin
   dx = a*x - b*x*y
-  dy = -c*y + d*x*y
-end a b c d
+  dy = -c*y + x*y
+end a b c
 
-p = [1.5,1.0,3.0,1.0]
+p = [1.5,1.0,3.0]
 prob = ODELocalSensitivityProblem(f,[1.0;1.0],(0.0,10.0),p)
 ```
 
@@ -135,8 +135,7 @@ autodifferentiation and numerical differentiation through the ODE solver:
 ```julia
 using ForwardDiff, Calculus
 function test_f(p)
-  pf = ParameterizedFunction(f,p)
-  prob = ODEProblem(pf,eltype(p).([1.0,1.0]),eltype(p).((0.0,10.0)))
+  prob = ODEProblem(f,eltype(p).([1.0,1.0]),eltype(p).((0.0,10.0)),p)
   solve(prob,Vern9(),abstol=1e-14,reltol=1e-14,save_everystep=false)[end]
 end
 

docs/src/types/bvp_types.md

@@ -48,15 +48,6 @@ bc!(residual, sol, p)
 where `u` is the current solution to the ODE which is used to compute the `residual`.
 Note that all features of the `ODESolution` are present in this form.
 In both cases, the size of the residual matches the size of the initial condition.
-For more general problems, use the
-[parameter estimation routines](../../analysis/parameter_estimation.html) and change the signature of `bc!` and `f` to
-
-```julia
-bc!(residual, sol, p)
-f(t, u, p) # or inplace as f(t, u, p, du)
-```
-
-where `p`is the `Vector` of free parameters.
 
 ### Fields
 

docs/src/types/dynamical_types.md

@@ -6,7 +6,7 @@ These algorithms require a Partitioned ODE of the form:
 
 ```math
 \frac{du}{dt} = f_1(v) \\
-\frac{dv}{dt} = f_2(u,p,t) \\
+\frac{dv}{dt} = f_2(u,t) \\
 ```
 This is a Partitioned ODE partitioned into two groups, so the functions should be
 specified as `f1(du,u,v,p,t)` and `f2(dv,u,v,p,t)` (in the inplace form), where `f1`
@@ -105,7 +105,7 @@ HamiltonianProblem{T}(H,q0,p0,tspan;kwargs...)
 
 ### Fields
 
-* `H`: The Hamiltonian `H(q,p)` which returns a scalar.
+* `H`: The Hamiltonian `H(q,p,params)` which returns a scalar.
 * `q0`: The initial positions.
 * `p0`: The initial momentums.
 * `tspan`: The timespan for the problem.