Skip to content

Instantly share code, notes, and snippets.

@IainNZ
Created October 20, 2014 03:45
Show Gist options
  • Save IainNZ/aa75b3298d0ad6a70d86 to your computer and use it in GitHub Desktop.
Save IainNZ/aa75b3298d0ad6a70d86 to your computer and use it in GitHub Desktop.
N20, R50, MIPGap=0.01, Threads=1
PYTHON
2651
-75.3212163
49.8698305592
min 7.85425305367
mean 7.98054280281
75656742.8543
cb_work_time/ITERS 1.93139100075e-06
JULIA
getObjectiveValue(m) => 75.32122134681227
norm((getValue(x))[:]) => 49.86983055916673
callbacks => 2651
minimum(times) => 7.035883033
median(times) => 7.1289896185
mean(times) => 7.1385828574
dummy => 6.5e8
cb_work_time / ITERS => 2.8016772899999997e-7
CB math in Python in 10x slower than Julia
Python est math time = 2651*1.93139100075e-06 = 0.00512011754298825
Julia est math time = 2651*2.801677289999e-7 = 0.00074272464957899
Therefore math time is not to blame for difference
Therefore 12% penalty just from Python interface overhead (extracting solution, adding constraints)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment