Merge pull request #150 from JuliaOpt/release-0.4

Release 0.4

.travis.yml

@@ -3,6 +3,7 @@ os:
   - linux
 julia:
   - 0.5
+  - 0.6
 notifications:
   email: false
 

README.md

@@ -4,7 +4,7 @@
 **WARNING:** *This package is currently in development. Any help or feedback is appreciated.*
 
 
-**Latest release:** 0.3.0
+**Latest release:** 0.4.0
 
 | **Documentation** | **Build Status** | **Social** |
 |:-----------------:|:----------------:|:----------:|
@@ -28,13 +28,13 @@ It is built upon [JuMP]
 - Linear dynamics
 - Linear or convex piecewise linear cost
 
-Extension to non-linear formulation are under development. 
+Extension to non-linear formulation are under development.
 Extension to more complex alea dependance are under developpment.
 
 ## Why Extensive formulation ?
 
 An extensive formulation approach consists in representing the stochastic problem as a deterministic
-one with more variable and call a standard deterministic solver. Mainly usable in a linear 
+one with more variable and call a standard deterministic solver. Mainly usable in a linear
 setting. Computational complexity is exponential in the number of stages.
 
 ## Why Stochastic Dynamic Programming ?
@@ -54,7 +54,10 @@ control strategies.
 
 
 ## Installation
-Installing StochDynamicProgramming is an easy process. Open Julia and enter
+Installing StochDynamicProgramming is an easy process.
+Currently, the package depends upon `StochasticDualDynamicProgramming.jl`, which is not
+yet registered in Julia's METADATA. To install the package,
+open Julia and enter
 
 ```julia
 julia> Pkg.update()

REQUIRE

@@ -1,6 +1,7 @@
 julia 0.5
-JuMP 0.16
+JuMP 0.17
 Distributions
 ProgressMeter
 Interpolations
-Iterators
+CutPruners 0.0.2 
+Compat 0.18

doc/sddp_api.rst

@@ -20,8 +20,8 @@ but allows to use either piecewise linear or quadratic costs.
 To define a `LinearSPModel`, the constructor is::
 
     spmodel = LinearSPModel(
-                        nstage,             # number of stages
-                        ubounds,            # bounds of control
+                        n_stage,             # number of stages
+                        u_bounds,            # bounds of control
                         x0,                 # initial state
                         cost,               # cost function
                         dynamic,            # dynamic
@@ -36,8 +36,8 @@ Default parameters
 ^^^^^^^^^^^^^^^^^^
 You should at least specify these parameters to define a `LinearSPModel`:
 
-- `nstage` (Int): number of stages in the stochastic multistage problem
-- `ubounds` (list of tuple): bounds upon control, defined as a sequence of tuple :code:`(umin, umax)`.
+- `n_stage` (Int): number of stages in the stochastic multistage problem
+- `u_bounds` (list of tuple): bounds upon control, defined as a sequence of tuple :code:`(umin, umax)`.
 - `x0` (`Vec{Float64}`): initial state
 - `cost` (`Function`): cost function as a function of time, state, control and noise returning a Float
 - `dynamic` (`Function`): system's dynamic as a function of time, state, control and noise returning a vector

examples/battery_storage_parallel.jl

@@ -106,10 +106,10 @@ println("library loaded")
                                                     controlSteps, infoStruct)
 end
 
-Vs = StochDynamicProgramming.solve_DP(spmodel,paramSDP, 1)
+Vs = StochDynamicProgramming.solve_dp(spmodel,paramSDP, 1)
 
 lb_sdp = StochDynamicProgramming.get_bellman_value(spmodel,paramSDP,Vs)
 println("Value obtained by SDP: "*string(lb_sdp))
-costsdp, states, stocks = StochDynamicProgramming.sdp_forward_simulation(spmodel,paramSDP,scenarios,Vs)
+costsdp, states, stocks = StochDynamicProgramming.forward_simulations(spmodel,paramSDP,Vs,scenarios)
 println(mean(costsdp))