Values are fold improvements: >1 means ncreduce is faster, <1 means ncreduce is slower
Columns are different array sizes: (4x4) (40x4) (4000x10) (10x4000) (4000x10000)
Rows are different types of reduce operation:
 A.f()
 A.T.f()
 A.f(0)
 A.f(1)
 A.T.f(1)
 A[2].f()
 A[:,2].f()


For SUM function
[[  3.8070922    3.74078624   3.18315868   3.21160784   2.43980108]
 [  4.46617647   4.22613065   8.31738401   6.8606588   29.21819473]
 [  1.9447132    1.56371814   0.96617167   3.01338318   7.0270556 ]
 [  2.51067616   4.4812749    6.82775021   2.81078355   2.02944134]
 [  1.91545392   1.62806468   0.87334081   2.54212652   0.80367664]
 [  3.6557377    3.52655539   3.45639535   3.09458466   3.09918836]
 [  3.20726783   2.93624557   1.0432128    3.20680628   0.96305311]]

For MEAN function
[[  4.53997379   4.33751425   3.18691456   3.21325964   2.61187721]
 [  5.09466667   4.86712486   8.09895611   6.66878963  29.56070697]
 [  1.32862454   1.2232241    0.95682217   2.77428247   6.59373701]
 [  1.38306992   2.07929176   5.6042247    2.67118025   2.01389827]
 [  1.32216749   1.24022472   0.8609297    2.38601915   0.79781743]
 [  4.26244953   4.21212121   4.28808864   3.22779923   3.19200338]
 [  4.05635649   3.68317972   1.08496276   4.00253807   0.95483616]]

For STD function
[[ 16.37931034  13.23200859   8.515398     8.6650662    5.56770624]
 [ 17.59516616  13.92785794  11.48111037  10.94399828  16.72880933]
 [  7.96362619   4.69929338   2.06533419   2.63442178   5.35333711]
 [ 12.23589744  12.1507772    8.82073033   8.74585517   6.89534047]
 [  7.9735658    4.66921665   1.54624704   2.49588909   1.30503195]
 [ 17.3400936   17.64123377  17.10280374   5.49334051   4.91767663]
 [ 14.07474227  10.96748768   1.73907847  13.80691299   1.08550626]]

For VAR function
[[ 12.75496689  11.19303797   8.23840394   8.17942839   5.35515979]
 [ 14.72340426  11.67340426  11.47882782  11.10354571  16.79689987]
 [  7.43418014   4.3738514    2.07330643   2.68918206   5.09448361]
 [ 11.42946429  13.17880795   9.98134987   8.36491556   6.79879305]
 [  7.52822581   4.28857407   1.4730624    2.47868726   1.1964753 ]
 [ 14.93135725  14.64387917  14.28         5.2311021    4.79074167]
 [ 12.18278146   9.37636544   1.65813694  11.40481928   1.04396634]]

For MAX function
[[  4.18282548   4.1835206    3.064325     3.09957884   1.89579732]
 [  5.3807267    5.09299191   8.16320492   6.81480951  23.32648022]
 [  2.22292994   1.53729977   0.82438672   1.95009491   6.05179257]
 [  2.95664467   4.09146341   3.55795579   2.05772538   1.74708854]
 [  2.35276753   1.66998507   0.86066282   1.75172327   0.80780005]
 [  4.53484603   4.39833333   3.88140556   3.10858712   3.08567761]
 [  4.31647399   3.50316857   1.21125174   3.87866109   0.93919081]]

For MIN function
[[  5.12647059   4.29441624   3.03089161   3.05712249   1.89781197]
 [  5.23816794   4.79036458   8.18441573   6.75257879  23.33387201]
 [  2.20617111   1.57129757   0.86588154   2.0558663    6.3874678 ]
 [  2.96841155   3.84391192   3.35497735   1.99594036   1.74332781]
 [  2.25244073   1.69904857   0.87264242   1.82947976   0.80815663]
 [  4.62801932   4.39569536   4.34477124   3.12006903   3.10487051]
 [  3.97965116   3.51935081   1.16092389   4.00438596   0.93541891]]

For ALL function
[[  2.79002467e-02   3.76257310e+00   9.28888889e+01   8.84859813e+01
    8.42846792e+04]
 [  4.99675325e+00   5.28160000e+00   1.31449298e+02   1.89387255e+02
    5.50127171e+05]
 [  2.30509745e+00   1.61616675e+00   3.98208860e-01   2.52603138e+00
    2.41908320e+00]
 [  2.82739726e+00   4.85934820e+00   3.36824142e+01   5.78263158e+01
    1.46416294e+03]
 [  1.95994914e+00   1.75441501e+00   4.47726973e-01   2.20169301e+00
    9.43527633e-01]
 [  4.18080000e+00   4.11092715e+00   4.09602649e+00   1.30212072e+01
    2.61726974e+01]
 [  3.84461538e+00   3.96411856e+00   1.29984496e+01   3.94307692e+00
    5.45761689e+01]]

For ANY function
[[  4.54672897e+00   4.97651007e+00   9.48595944e+01   9.28051948e+01
    8.83522646e+04]
 [  5.02138158e+00   5.54304636e+00   1.40057190e+02   1.87542130e+02
    5.78053539e+05]
 [  2.42564910e+00   1.70693277e+00   4.13649990e-01   2.52723749e+00
    2.52861281e+00]
 [  2.97480620e+00   5.02946429e+00   3.39985881e+01   5.69674115e+01
    1.50428346e+03]
 [  2.29548872e+00   1.81257078e+00   4.52254796e-01   2.20091677e+00
    9.46904281e-01]
 [  4.29111842e+00   4.16582064e+00   4.13087248e+00   1.33412162e+01
    2.89243697e+01]
 [  3.91051805e+00   3.96214511e+00   1.41828299e+01   3.93322734e+00
    6.09200000e+01]]

python profile.py 10948.30s user 625.08s system 99% cpu 3:13:06.33 total