jgillis

import casadi.*

% In this example, we fit a nonlinear model to measurements
%
% This example uses more advanced constructs than the vdp* examples:
% Since the number of control intervals is potentially very large here,
% we use memory-efficient Map and MapAccum, in combination with
% codegeneration.
%
% We will be working with a 2-norm objective:

jgillis

#!/usr/bin/python

import sys
import hashlib
import subprocess
import shutil
import os
args = sys.argv[1:]

cachedir = ".cache"

jgillis

casadi2matlabfun =@(F,castfun) @(varargin) subsref(cellfun(castfun,F(varargin),'UniformOutput',false),struct('type','{}','subs',{{':'}}));

Hf = casadi2matlabfun(H,@full)

jgillis

function [out] = casadi_struct2vec(s)
  flat = {};
  if isstruct(s)
    for f=fieldnames(s)'
      flat = {flat{:} casadi_struct2vec(s.(f{1}))};
    end
    out = vertcat(flat{:});
  else if iscell(s)
    for i=1:length(s)
       flat = {flat{:} casadi_struct2vec(s{i})};

jgillis

classdef PSDchecker < casadi.Callback2
  properties
      n
      fwd
      adj
  end
  methods
    function self = PSDchecker(n)
      self.n = n;
      self.fwd = mydergen(n,true);

jgillis

from functools import partial

class Infix(object):
    def __init__(self, func):
        self.func = func
    def __or__(self, other):
        return self.func(other)
    def __ror__(self, other):
        return Infix(partial(self.func, other))
    def __call__(self, v1, v2):

jgillis

# Pendulum collocation
# w position of cart
# w1,y1 position of pendulum

import numpy as NP
from casadi import *
from casadi.tools import *
import matplotlib.pyplot as plt

jgillis

function ret = classify_linear(e,v)
 %
 %   Takes vector expression e, and symbolic primitives v
 %   Returns classification vector
 %   For each element in e, determines if:
 %     - element is nonlinear in v               (2)
 %     - element is    linear in v               (1)
 %     - element does not depend on v at all     (0)
 %     
 %   This method can be sped up a lot with JacSparsityTraits::sp

jgillis

from casadi import *

H = DM([[1,-1],[-1,2]])
G = DM([-2,-6])
A =  DM([[1, 1],[-1, 2],[2, 1]])

LBA = DM([-inf]*3)
UBA = DM([2, 2, 3])

LBX = DM([0]*2)

jgillis

classdef Function < casadi.SharedObject
    %FUNCTION General function.
    %
    %
    %
    %A general function $f$ in casadi can be multi-input, multi-output. Number of
    %inputs: nin n_in() Number of outputs: nout n_out()  We can view this
    %function as a being composed of a ( nin, nout) grid of single-input, single-
    %output primitive functions. Each such primitive function $f_ {i, j}
    %\\forall i \\in [0, nin-1], j \\in [0, nout-1]$ can map as $\\mathbf