Skip to content

scipy/legacy-optimize

Rule legacy-optimize
Pack SciPy
Severity info
Fix none, edit by hand
Library versions every version

SciPy lists these optimize functions as legacy and not recommended for new scripts. Use the call that the message names, which returns an OptimizeResult.

xopt = optimize.fmin_bfgs(f, x0, fprime=grad, disp=False)
xopt = optimize.minimize(f, x0, jac=grad, method="BFGS").x

The rule reports the functions that scipy.optimize lists as legacy and not recommended for new scripts. Its message names the call that replaces each one:

Legacy function Write instead
fmin, fmin_powell, fmin_cg, fmin_bfgs, fmin_ncg, fmin_l_bfgs_b, fmin_tnc, fmin_cobyla, fmin_slsqp minimize(fun, x0, method=...) with 'Nelder-Mead', 'Powell', 'CG', 'BFGS', 'Newton-CG', 'L-BFGS-B', 'TNC', 'COBYLA' or 'SLSQP'
fminbound minimize_scalar(fun, bounds=(a, b), method='bounded')
brent, golden minimize_scalar(fun, method='brent') or method='golden'
leastsq least_squares(fun, x0)
fsolve root(fun, x0)
broyden1, broyden2, anderson, excitingmixing, linearmixing, diagbroyden root(fun, x0, method=...) with the same name
newton_krylov root(fun, x0, method='krylov')

xpd has no automatic fix. Because the newer functions take other argument names and return an OptimizeResult. Make the change by hand.

xpd prints this message with the issue. A name that starts with $ stands for code from your file.

SciPy lists optimize.$F among the legacy functions that are not recommended for new scripts. Write optimize.$NEW, which returns an OptimizeResult.

Browse the other rules of the SciPy pack.