Skip to content

@jgillis jgillis released this Oct 3, 2019

Install

Grab a binary from the table (for MATLAB, use the newest compatible version below):

WindowsLinuxMac
Matlab R2016a or later,
R2014b,
R2014a,
R2013a or R2013b
R2014b or later,
R2014a
R2015a or later,
R2014b,
R2014a
Octave 4.4.1 (32bit / 64bit),
4.4.0 (32bit / 64bit),
5.1.0 (32bit / 64bit)
4.4.1,
4.2.2,
5.1.0
5.1.0
Python Py27 (32bit* / 64bit*),
Py35 (32bit* / 64bit*),
Py36 (32bit* / 64bit*),
Py37 (32bit* / 64bit*)
Py27,
Py35,
Py36,
Py37
Py27,
Py35,
Py36,
Py37
or just pip install casadi (needs pip -V>=8.1)

(*) Check your Python console if you need 32bit or 64bit - bitness should be printed at startup.

Unzip in your home directory and adapt the path:

Matlab/OctavePython

addpath('<yourpath>/casadi-matlabR2014a-v3.4.5')
import casadi.*
x = MX.sym('x')
disp(jacobian(sin(x),x))


from sys import path
path.append(r"<yourpath>/casadi-py27-v3.4.5")
from casadi import *
x = MX.sym("x")
print(jacobian(sin(x),x))

Get started with the example pack.

Troubleshooting

Release notes

CasADi Functions

  • CasADi Functions can be serialized now (#308).
f.save('f.casadi')            % Dump any CasADi Function to a file
f = Function.load('f.casadi') % Loads back in

This enables easy sharing of models/solver isntances beteen Matlab/Python/C++ cross-platform, and enables a form of parallelization.

  • You can now evaluate CasADi Functions from C without requiring code-generation. This makes it possible to embed CasADi computations in Fortran, Julia, FMI, ...
  • All CasADi Functions support timing information now (print_time, default true for QP and NLP solvers). Use record_time to make timings available through f.stats() without printing them.
  • map with reduce arguments now has an efficient implementation (no copying/repmat)
  • Low-overhead Callback eval support was changed to eval_buffer
  • FunctionInternal::finalize no longer takes options dict.
  • Options always_inline and never_inline were added
  • Options is_diff_in and is_diff_out were added

CasADi expressions

  • breaking: IM type is removed from public API (was used to represent integer sparse matrices). Use DM instead.
  • breaking: linspace(0,1,3) and linspace(0.0,1,3) now both return [0 0.5 1] instead of [0 0 1] for the former
  • MX supports slicing with MX now (symbolic indexing).
  • Issue #2364:
    • breaking: veccat of an empty list now returns 0-by-1 instead of 0-by-0.
    • jtimes output dimensions have changed when any of the arguments is empty.
    • NLP function object's 'lam_p' is now 0-by-1 in case of missing parameters.

Interpolation functionality

  • Support for parametric (=changeable only, but not differentiable) grid and/or coefficients for linear/spline interpolation
    • for interpolant, new constructors where added that takes dimensions instead of concrete vectors
  • Support for symbolic (differentiable) grid and coefficients for linear interpolation (set inline option to true).

Python specific

  • Overhead-less CasADi Function evaluation API added through Python memoryviews
  • Similar functionality in Callbacks

Matlab/Octave specific

  • breaking: a(:)=b now behaves like Matlab builtin matrices when a is a CasADi matrix. Before, only the first column of a would be touched by this statement. (#2363)
  • breaking: Fixed bug where MX constructor treated a numeric row vector as column vector. Now size(MX(ones(1,4))) returns (1,4) as expected. (#2366)
  • Can now use spy directly on DM,MX,SX

Opti

  • Opti supports conic problems now: Opti('conic')
  • One can now easily obtain a parametric solution as a CasADi Function from an Opti instance:
opti = Opti()
x = opti.variable()
y = opti.variable()
p = opti.parameter()
      
opti.minimize(y**2+sin(x-y-p)**2)
opti.subject_to(x+y>=1)
      
opti.solver(nlpsolver,nlpsolver_options)

F = opti.to_function("F",[x,p,opti.lam_g],[x,y])
      
r = F(0,0.1,0)
  • Using Opti together with max_iter is more natural now: use solve_limited() to avoid exceptions to be raised when iterations or time runs out. No need to try/catch.

Code-generation

  • breaking: external now looks for a .dylib file, not .so
  • breaking: Codegen API has changes related to thread-local memory:
  • void* mem changed to int mem
  • alloc_mem, init_mem, free_mem have been purged. checkout and release replace them.
  int mem = checkout();
  eval(arg, res, iw, w, mem);
  release(mem);
  • Codegen 'CODEGEN_PREFIX' has been renamed to 'CASADI_CODEGEN_PREFIX'
  • QP solvers (QRQP, OSQP) and SqpMethod codegenerate now. This means that embedded MPC with CasADi is now more practical.
  • Runge-Kutta and Collocation Integrator objects can be inlined into code-generatable MX Function with the 'simplify' option.

Solvers

  • breaking: NLP solvers - bound_consistency, an option to post-process the primal and dual solution by projecting it on the bounds, introduced in 3.4, was changed to default off

  • Sundials was patched to support multi-threading

  • WORHP was bumped to v1.13

  • SNOPT was bumped to v7.7

  • SuperSCS (conic solver) was added

  • OSQP (QP solver) was added

  • CBC (LP solver) was added

  • breaking: SQP Method

    • regularize_margin option was added
    • regularize (bool) option was removed. To get the effect of regularize=true, specify convexify_strategy='regularize'. Other strategies include clipping eigenvalues.
    • line search was reverted from a custom designed thing, to standard textbook L1
  • CPLEX and Gurobi got support for sos constraints

  • Conic/qpsol interface extended for semidefinite programming and SOCP

    • Solvers supporting SOCP: Gurobi, SuperSCS, CPLEX
  • breaking: Newton Rootfinder now supports a line_search option (default true)

  • Rootfinder now throws an exception by default ('error_on_fail' option true) when failing to converge

Convenience tools

  • Debugging facilities:
    • Function options print_in/print_in print inputs/outputs when numerically evaluating a function
    • Function option dump_in/dump_out dumps to the file system
    • Function option dump dumps the function itself (loadable with Function.load)
  • DM.from_file and DM.to_file with a MatrixMarket and txt support
  • Helping interaction with codegen with main=true: Function.generate_in/Function.nz_from_in/Function.nz_to_in to help creating input text files.
  • Function.convert_in/Function.convert_out to switch between list and dictionary arguments/results
Assets 47
README.md 81 Bytes
You can’t perform that action at this time.