1、基本数组统计方法
基本都可以使用np.sum(a,axis = 0)和a.sum(0)两种格式
np.sum(a,axis = 0):求和
np.mean(a,axis=0):算数平均数
np.std(a,axis = 0),np.var(a,axis = 0):标准差和方差
# 标准差也被称为标准偏差,或者实验标准差,公式为
#方差是实际值与期望值之差平方的平均值,而标准差是方差算术平方根。
np.min(a,axis = 0),np.max(a,axis = 0):最大值和最小值
np.argmin(a,axis=0) , np.argmax(a,axis = 0):最大值和最小值的索引
np.cumsum(a,axis=0), np.cumprod(a,axis = 0):累计求和、累计求积(结果直接展平为一维数组)
a
array([[0, 1, 2],
[3, 4, 5],
[6, 7, 8]])
>>> a.sum()
36
>>> np.sum(a)
36
>>> a.mean(1)
array([ 1., 4., 7.])
>>> a.mean(0)
array([ 3., 4., 5.])
>>> a.mean(axis = 1)
array([ 1., 4., 7.])
>>> a.mean(axis = 0)
array([ 3., 4., 5.])
>>> np.mean(a,axis = 1)
array([ 1., 4., 7.])
>>> np.mean(a,axis = 0)
array([ 3., 4., 5.])
>>> np.mean(a,0)
array([ 3., 4., 5.])
>>> np.std(a,0)
array([ 2.44948974, 2.44948974, 2.44948974])
>>> np.std(a,1)
array([ 0.81649658, 0.81649658, 0.81649658])
>>> np.var(a)
6.666666666666667
>>> np.var(a,0)
array([ 6., 6., 6.])
>>> 6**0.5
2.449489742783178
>>> (2/3)**0.5
0.816496580927726
>>> a.var(1)
array([ 0.66666667, 0.66666667, 0.66666667])
>>> a.min()
0
>>> a.max()
8
>>> a.min(1)
array([0, 3, 6])
>>> a.argmin()
0
>>> a
array([[0, 1, 2],
[3, 4, 5],
[6, 7, 8]])
>>> a.argmax()
8
>>> a[[0,2,1]]
array([[0, 1, 2],
[6, 7, 8],
[3, 4, 5]])
>>> a[[0,2,1]].argmax()
5
>>> a.argmax(0)
array([2, 2, 2], dtype=int64)
>>> a.argmin(1)
array([0, 0, 0], dtype=int64)
>>> a
array([[0, 1, 2],
[3, 4, 5],
[6, 7, 8]])
>>> a.cumsum()
array([ 0, 1, 3, 6, 10, 15, 21, 28, 36], dtype=int32)
>>> a.cumprod()
array([0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=int32)
>>> a.sum(0)
array([ 9, 12, 15])
>>> np.cumprod(a,0)
array([[ 0, 1, 2],
[ 0, 4, 10],
[ 0, 28, 80]], dtype=int32)