Skip to content

Instantly share code, notes, and snippets.

@jeetsukumaran

jeetsukumaran/results.tsv

Last active April 21, 2018 03:42
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save jeetsukumaran/f9d95af1c4833e4d183aa5464b9529f8 to your computer and use it in GitHub Desktop.
Save jeetsukumaran/f9d95af1c4833e4d183aa5464b9529f8 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
results.tsv
L-BFGS-B 0.0 0.000000 True: b'CONVERGENCE: NORM_OF_PROJECTED_GRADIENT_<=_PGTOL'
L-BFGS-B 0.1 0.100000 True: b'CONVERGENCE: NORM_OF_PROJECTED_GRADIENT_<=_PGTOL'
L-BFGS-B 0.2 0.200000 True: b'CONVERGENCE: NORM_OF_PROJECTED_GRADIENT_<=_PGTOL'
L-BFGS-B 0.3 0.300000 True: b'CONVERGENCE: NORM_OF_PROJECTED_GRADIENT_<=_PGTOL'
L-BFGS-B 0.4 0.400000 True: b'CONVERGENCE: NORM_OF_PROJECTED_GRADIENT_<=_PGTOL'
L-BFGS-B 0.5 0.500000 True: b'CONVERGENCE: NORM_OF_PROJECTED_GRADIENT_<=_PGTOL'
L-BFGS-B 0.6 0.500000 True: b'CONVERGENCE: REL_REDUCTION_OF_F_<=_FACTR*EPSMCH'
L-BFGS-B 0.7 0.500000 True: b'CONVERGENCE: REL_REDUCTION_OF_F_<=_FACTR*EPSMCH'
L-BFGS-B 0.8 0.500000 True: b'CONVERGENCE: REL_REDUCTION_OF_F_<=_FACTR*EPSMCH'
L-BFGS-B 0.9 0.656250 True: b'CONVERGENCE: REL_REDUCTION_OF_F_<=_FACTR*EPSMCH'
L-BFGS-B 1.0 0.519531 False: b'ABNORMAL_TERMINATION_IN_LNSRCH'
TNC 0.0 0.000000 True: Local minimum reached (|pg| ~= 0)
TNC 0.1 0.100000 True: Converged (|f_n-f_(n-1)| ~= 0)
TNC 0.2 0.200002 True: Converged (|f_n-f_(n-1)| ~= 0)
TNC 0.3 0.300000 True: Converged (|f_n-f_(n-1)| ~= 0)
TNC 0.4 0.400000 True: Converged (|f_n-f_(n-1)| ~= 0)
TNC 0.5 0.500000 True: Local minimum reached (|pg| ~= 0)
TNC 0.6 0.600000 True: Converged (|f_n-f_(n-1)| ~= 0)
TNC 0.7 0.700000 True: Converged (|f_n-f_(n-1)| ~= 0)
TNC 0.8 0.500000 False: Linear search failed
TNC 0.9 0.500000 False: Linear search failed
TNC 1.0 0.500000 False: Linear search failed
SLSQP 0.0 0.000000 True: Optimization terminated successfully.
SLSQP 0.1 0.099992 True: Optimization terminated successfully.
SLSQP 0.2 0.200135 True: Optimization terminated successfully.
SLSQP 0.3 0.300000 True: Optimization terminated successfully.
SLSQP 0.4 0.399910 True: Optimization terminated successfully.
SLSQP 0.5 0.500000 True: Optimization terminated successfully.
SLSQP 0.6 0.600090 True: Optimization terminated successfully.
SLSQP 0.7 0.700000 True: Optimization terminated successfully.
SLSQP 0.8 0.799865 True: Optimization terminated successfully.
SLSQP 0.9 0.900008 True: Optimization terminated successfully.
SLSQP 1.0 nan False: Iteration limit exceeded
Nelder-Mead 0.0 0.000012 True: Optimization terminated successfully.
Nelder-Mead 0.1 0.100000 True: Optimization terminated successfully.
Nelder-Mead 0.2 0.200000 True: Optimization terminated successfully.
Nelder-Mead 0.3 0.300000 True: Optimization terminated successfully.
Nelder-Mead 0.4 0.400000 True: Optimization terminated successfully.
Nelder-Mead 0.5 0.500000 True: Optimization terminated successfully.
Nelder-Mead 0.6 0.600000 True: Optimization terminated successfully.
Nelder-Mead 0.7 0.700000 True: Optimization terminated successfully.
Nelder-Mead 0.8 0.800000 True: Optimization terminated successfully.
Nelder-Mead 0.9 0.900000 True: Optimization terminated successfully.
Nelder-Mead 1.0 0.999988 True: Optimization terminated successfully.
Powell 0.0 199.770530 False: Maximum number of function evaluations has been exceeded.
Powell 0.1 199.770530 False: Maximum number of function evaluations has been exceeded.
Powell 0.2 199.770530 False: Maximum number of function evaluations has been exceeded.
Powell 0.3 199.770530 False: Maximum number of function evaluations has been exceeded.
Powell 0.4 199.770530 False: Maximum number of function evaluations has been exceeded.
Powell 0.5 199.770530 False: Maximum number of function evaluations has been exceeded.
Powell 0.6 199.770530 False: Maximum number of function evaluations has been exceeded.
Powell 0.7 199.770530 False: Maximum number of function evaluations has been exceeded.
Powell 0.8 199.770530 False: Maximum number of function evaluations has been exceeded.
Powell 0.9 199.770530 False: Maximum number of function evaluations has been exceeded.
Powell 1.0 199.770530 False: Maximum number of function evaluations has been exceeded.
BFGS 0.0 -19.499997 False: Desired error not necessarily achieved due to precision loss.
BFGS 0.1 -159.499984 False: Desired error not necessarily achieved due to precision loss.
BFGS 0.2 -539.499965 False: Desired error not necessarily achieved due to precision loss.
BFGS 0.3 -959.499982 False: Desired error not necessarily achieved due to precision loss.
BFGS 0.4 -839.500036 False: Desired error not necessarily achieved due to precision loss.
BFGS 0.5 0.500000 True: Optimization terminated successfully.
BFGS 0.6 840.499964 False: Desired error not necessarily achieved due to precision loss.
BFGS 0.7 960.500018 False: Desired error not necessarily achieved due to precision loss.
BFGS 0.8 540.500035 False: Desired error not necessarily achieved due to precision loss.
BFGS 0.9 160.500017 False: Desired error not necessarily achieved due to precision loss.
BFGS 1.0 20.500003 False: Desired error not necessarily achieved due to precision loss.
Raw
scipy_optimization_exp.py
#! /usr/bin/env python
# -*- coding: utf-8 -*-
import numpy as np
from scipy.optimize import minimize
from scipy.stats import binom
def main():
n = 10
x0 = np.array([0.5])
bounded_methods = ("L-BFGS-B", "TNC", "SLSQP")
unbounded_methods = ("Nelder-Mead", "Powell", "BFGS", "COBYLA", )
for method in bounded_methods + unbounded_methods:
for k in range(0,11):
def f(x, *args):
return -1 * binom.pmf(k, n, x[0])
kwargs = {"method": method}
if method in bounded_methods:
kwargs["bounds"] = ((0.0, 1.0),)
res = minimize(f, x0, **kwargs)
try:
est = res.x[0]
except IndexError:
est = res.x
print("{:10}\t{:0.1f}\t{:8.6f}\t{}: {}".format(method, k/float(n), est, res.success, res.message))
if __name__ == '__main__':
main()
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment