numpy3.统计相关

2024-04-04 22:38:54

numpy3.统计相关

次序统计
- 最小值
- 最大值
- 极差
- 分位数
均值与方差
相关

次序统计

最小值

np.amin(a, axis=None, out=None, keepdims=np._NoValue, initial=np._NoValue,where=np._NoValue)

最大值

np.amax(a, axis=None, out=None, keepdims=np._NoValue, initial=np._NoValue,where=np._NoValue)

极差

peak to peak
numpy.ptp(a, axis=None, out=None, keepdims=np._NoValue)

分位数

a为array，q为分位数值(0-100)，q可以为数组此时取多个分位数，axis范围0或1。返回第q%小的数。
np.percentile(a, q, axis=None, out=None, overwrite_input=False,interpolation='linear', keepdims=False)

均值与方差

中位数

np.median(a, axis=None, out=None, overwrite_input=False, keepdims=False)

平均值

np.mean(a, axis=None, dtype=None, out=None, keepdims=np._NoValue))

加权平均值

weights为权值数组。
np.average(a, axis=None, weights=None, returned=False)

方差

方差分母为n，样本方差无偏估计分母为n-1
ddof：*度的个数。
np.var(a, axis=None, dtype=None, out=None, ddof=0, keepdims=np._NoValue)

标准差

np.std(a, axis=None, dtype=None, out=None, ddof=0, keepdims=np._NoValue)

相关

协方差矩阵

协方差直观含义为两个变量总体误差的期望，描述两个变量协同变化的程度：
Cov(X, Y)
= E[(X-E(X)) * (Y-E(Y))]
= E[XY] - 2E[Y] * E[X] + E[X] * E[Y]
= E[XY] - E[X] * E[Y]
Cov(X, X) 即得到 X 的方差
下方函数只给出m计算方差，给出m和y计算协方差矩阵(对称)：
np.cov(m, y=None, rowvar=True, bias=False, ddof=None, fweights=None,aweights=None)

直方图

bins：一维单调数组，有序排列，代表横坐标。right：间隔是否包含最右。返回x在bin中的位置。
np.digitize(x, bins, right=False)

我是代码：

import numpy as np

np.random.seed(926734542)

x = np.random.randint(0, 9, 9).reshape((3, 3))
print(x)

xmin = np.amin(x, axis=0)
print(xmin)

xmax = np.amax(x, axis=1)
print(xmax)

peak = np.ptp(x, axis=0)
print(peak)

perc = np.percentile(x, 50, axis=0)
print(perc)

medi = np.median(x, axis=0)
print(medi)

mean = np.mean(x, axis=0)
print(mean)

aver = np.average(x, axis=0, weights=[0, 1, 2])
print(aver)

var = np.var(x)
print(var)

std = np.std(x)
print(std)

np.random.seed(5426)
y = np.random.randint(0, 9, 9).reshape((3, 3))
cov = np.cov(x, y)
print(cov)

corr = np.corrcoef(x, y)
print(corr)

z = np.arange(0, 9)
digi = np.digitize(x, z, True)
print(digi)

码农公寓

numpy3.统计相关