data science学习笔记1

Mutiple Plots on One Graph
plt.plot(x, norm.pdf(x))
plt.plot(x, norm.pdf(x, 1.0, 0.2)) #1.0 = mean, 0.2 = DS
plt.show()

使用plt.savefig 所保存图片为空白：

解决方法：在plt.show()之前调用plt.savefig

画散点图

from pylab import randn
X = randn(10000)
Y = randn(10000)
plt.scatter(X,Y) #注意顺序，先画图再添加坐标轴
axes = plt.axes()
axes.set_xlim([0, 1])
axes.set_ylim([0, 4])
plt.show()

covariance：协方差

协方差>0:x,y同向变化，且协方差越大同向程度越高

协方差<0:x,y反向变化，且协方差绝对值越大反向程度越高

correlation计算：covariance/SD

-1:perfect inverse correlation,0:no correlation,1:perfect correlation

贝叶斯公式

#字典，计算不同年龄段人群购买数量
from numpy import random
random.seed(0)

totals = {20:0, 30:0, 40:0, 50:0, 60:0, 70:0}
purchases = {20:0, 30:0, 40:0, 50:0, 60:0, 70:0}
totalPurchases = 0
for _ in range(100000):
    age = random.choice([20, 30, 40, 50, 60, 70])
    purchaseProbability = float(age) / 100.0 #除法运算float
    totals[age] += 1
    if (random.random() < purchaseProbability):
        totalPurchases += 1
        purchases[age] += 1