python2 .x到python 3.x：'float'和'NoneType'实例之间不支持'>'

如何解决python2 .x到python 3.x：'float'和'NoneType'实例之间不支持'>'

在python 3.x版本中执行（Python 2.x代码）时发生以下错误：

" File "C:\Users\silen\Documents\♠ Asset Allocation
Project\HRP(Hierarchical Risk Parity)_py2.7\Python3.7\CLA.py",line
47,in solve
    if (self.l[-1] is None or l<self.l[-1]) and l>l_out:l_out,i_out=l,i
 TypeError: '>' not supported between instances of 'float' and 'NoneType' "

解决方案是什么？

class CLA:   
    def __init__(self,mean,covar,lB,uB):
        # Initialize the class
        self.mean=mean
        self.covar=covar
        self.lB=lB
        self.uB=uB
        self.w=[] # solution
        self.l=[] # lambdas
        self.g=[] # gammas
        self.f=[] # free weights



    def solve(self):
        # Compute the turning points,free sets and weights
        f,w=self.initAlgo()
        self.w.append(np.copy(w)) # store solution
        self.l.append(None)
        self.g.append(None)
        self.f.append(f[:])
        while True:
            #1) case a): Bound one free weight
            l_in=None
            if len(f)>1:
                covarF,covarFB,meanF,wB=self.getMatrices(f)
                covarF_inv=np.linalg.inv(covarF)
                j=0
                for i in f:
                    l,bi=self.computeLambda(covarF_inv,wB,j,[self.lB[i],self.uB[i]])
                    if l>l_in:l_in,i_in,bi_in=l,i,bi
                    j+=1
            #2) case b): Free one bounded weight
            l_out=None
            if len(f)<self.mean.shape[0]:
                b=self.getB(f)
                for i in b:
                    covarF,wB=self.getMatrices(f+[i])
                    covarF_inv=np.linalg.inv(covarF)
                    l,meanF.shape[0]-1,\
                        self.w[-1][i])
                    if (self.l[-1] is None or l<self.l[-1]) and l>l_out:l_out,i          
            if (l_in is None or l_in < 0) and (l_out is None or l_out < 0):
                #3) compute minimum variance solution
                self.l.append(0)
                covarF,wB=self.getMatrices(f)
                covarF_inv=np.linalg.inv(covarF)
                meanF=np.zeros(meanF.shape)
            else:
                #4) decide lambda
                if l_in>l_out:
                    self.l.append(l_in)
                    f.remove(i_in)
                    w[i_in]=bi_in # set value at the correct boundary
                else:
                    self.l.append(l_out)
                    f.append(i_out)
                covarF,wB=self.getMatrices(f)
                covarF_inv=np.linalg.inv(covarF)
            #5) compute solution vector
            wF,g=self.computeW(covarF_inv,wB)
            for i in range(len(f)):w[f[i]]=wF[i]
            self.w.append(np.copy(w)) # store solution
            self.g.append(g)
            self.f.append(f[:])
            if self.l[-1]==0:break
        #6) Purge turning points
        self.purgeNumErr(10e-10)
        self.purgeExcess()

    def getCLA(cov,**kargs):
        # Compute CLA's minimum variance portfolio
        mean = np.arange(cov.shape[0]).reshape(-1,1)
        # Not used by C portf
        lB = np.zeros(mean.shape)
        uB = np.ones(mean.shape)
        cla = CLA.CLA(mean,cov,uB)
        cla.solve()
        return cla.w[-1].flatten()

w_ = pd.Series(getCLA(cov=cov_,corr=corr_)) 

解决方法

您无法在Python 3中比较不同的类型。幸运的是，在这种情况下，看起来所有新的l都是float，而您只想将“小于所有”的值作为初始化程序，对于float with -inf来说很简单：

from math import inf

... 
l_in = -inf  # Instead of None