很多商界领袖，如李嘉诚，他们并没有很高的学历，是如何做到准确分析经济走势的？ 第1页

       import numpy as np import matplotlib.pyplot as plt  #parameters #true model coeff=np.random.uniform(-2,2,20) #education tell you only 6 out of 20 true parameters education=np.random.choice(range(20),6,replace=False)  #generate factor and true economic fundamental, 100 periods factor=np.random.normal(0,1,[20,100]) ytrue=np.dot(coeff,factor)  #educated guess of economic fundamental coeffedu=np.zeros(20) for edu in education:     coeffedu[edu]=coeff[edu] yedu=np.dot(coeffedu,factor)  #uneducated parameters, take -1, 0 or 1, a.k.a signs of parameter #people knows 1,5,10,15,or 20 signs of the true parameter, but not the magnitude def unedu(numfac):     returntemp=np.zeros(20)     choicetemp=np.random.choice(range(20),numfac,replace=False)     for cho in choicetemp:         if coeff[cho]!=0:             returntemp[cho]=coeff[cho]/np.abs(coeff[cho])     return returntemp unedu1=unedu(1) unedu5=unedu(5) unedu10=unedu(10) unedu15=unedu(15) unedu20=unedu(20)  #prediction from uneducated people yunedu1=np.dot(unedu1,factor) yunedu5=np.dot(unedu5,factor) yunedu10=np.dot(unedu10,factor) yunedu15=np.dot(unedu15,factor) yunedu20=np.dot(unedu20,factor)  #plot def plotpred(series,labelvalue):     fig,example=plt.subplots(figsize=(16,10))     example.plot(series,label=labelvalue)     example.plot(ytrue,label='True value of economic fundamental')     example.legend(loc='right')     example.set_title('Predicted VS true, educated VS uneducated')     example.set_xlabel('Time')     example.set_ylabel('Economic fundamental')     plt.show()  plotpred(yedu,'Educated guess') plotpred(yunedu1,'Uneducated guess, know 1 signs') plotpred(yunedu5,'Uneducated guess, know 5 signs') plotpred(yunedu10,'Uneducated guess, know 10 signs') plotpred(yunedu15,'Uneducated guess, know 15 signs') plotpred(yunedu20,'Uneducated guess, know 20 signs')