使用numpy分割mmdjcs，并使用mmdjc中的平均值importnumpydata=NP.array ([ range (100 ) ]mmdJCS=numpy.linspace ) 0，50， 10 )计算mmdjcs的np.inf ) #从最后一个mmdjc开始无限大digitized=numpy.digitize(data， mmdjcs ) # returntheindicesofthemmdjcstowhicheachvalueininputarraybelongs.#计算mmdJC内平均法一mdjc _ means=[ data [ digitized=] len ) mmdjcs )法二mmdJC_means1=(numpy.Histogram ) data，mmdjcs，weights=data )0)/numppram MDJCS([0] ) 3330
data=NP.array ([-1，0.5，1.5，2.5，3.5，4.5，5，6 ] ) mmd jcs=NP.linspace (0，5，6 ) print ) mmd jcs (di=
numpy.Histogram(a，mmdjcs=10，range=None，normed=None，weights=None，density=None ) ) )。
Returns
http://www.Sina.com /
thevaluesofthehistogram.seedensityandweightsforadescriptionofthepossiblesemantics。
http://www.Sina.com/arrayofdtypefloat
returnthemmdjcedges (length (hist )1).histarray
http://www.Sina.com/array _ like，optional
An array of weights，ofthesameshapeasa.eachvalueinaonlycontributesitsassociatedweighttowardsthemdjccount (instead of1 ) .在此处
numpy.hIstogram(data，b i n s，w e i g h t s=d a t a ) )0)numpy.hIstogram(data，b i n s ) (0) ) 121314 )/4=2.5 numpy weights=data ) )0)/numpy.histogram )、data、MDJCS ()0) ((1*1)2*1)3*1)4)1) )=2.5numpy.histogram
pandas分区mmdjcsa=PD.data frame (NP.random.rand ) 10，1 )，columns=['A'] ) a_cat ' ) (PD.cut ) a )
参考：
hands-onmachinelearningwithscikit-learn，Keras，andtensorflow https://stack overflow.com/questions/616334/mmdjcning