数据分析和科学计算可视化(6)

当前位置:
首页 > temp > 简明python教程 >

数据分析和科学计算可视化(6)
矩阵不同维度上的计算

>>> x = np.matrix(np.arange(0,10).reshape(2,5))  # 二维矩阵

>>> x

matrix([[0, 1, 2, 3, 4],

        [5, 6, 7, 8, 9]])

>>> x.sum()                                      # 所有元素之和

45

>>> x.sum(axis=0)                                # 纵向求和

matrix([[ 5,  7,  9, 11, 13]])

>>> x.sum(axis=1)                                # 横向求和

matrix([[10],

        [35]])

>>> x.mean()                                     # 平均值

4.5

>>> x.mean(axis=1)

matrix([[ 2.],

        [ 7.]])

>>> x.mean(axis=0)

matrix([[ 2.5,  3.5,  4.5,  5.5,  6.5]])

>>> x.max()                                # 所有元素最大值

9

>>> x.max(axis=0)                          # 纵向最大值

matrix([[5, 6, 7, 8, 9]])

>>> x.max(axis=1)                          # 横向最大值

matrix([[4],

        [9]])

>>> weight = [0.3, 0.7]                    # 权重

>>> np.average(x, axis=0, weights=weight)

matrix([[ 3.5,  4.5,  5.5,  6.5,  7.5]])

>>> x = np.matrix(np.random.randint(0, 10, size=(3,3)))

>>> x

matrix([[3, 7, 4],

        [5, 1, 8],

        [2, 7, 0]])

>>> x.std()                         # 标准差

2.6851213274654606

>>> x.std(axis=1)                   # 横向标准差

matrix([[ 1.69967317],

        [ 2.86744176],

        [ 2.94392029]])

>>> x.std(axis=0)                   # 纵向标准差

matrix([[ 1.24721913,  2.82842712,  3.26598632]])

>>> x.var(axis=0)                   # 纵向方差

matrix([[  1.55555556,   8.        ,  10.66666667]])
栏目列表