数据类型
import numpy as np
arr1 = np.array([1,2,3],dtype=np.float64)
arr1
array([1., 2., 3.])
arr2 = np.array([1,2,3],dtype=np.int32)
arr2
array([1, 2, 3], dtype=int32)
arr1.dtype
dtype('float64')
arr2.dtype
dtype('int32')
新手不必担心,用时再考虑即可
2.数据类型转换
arr = np.array([1,2,3,4,5])
arr.dtype
dtype('int64')
float_arr = arr.astype(np.float64) #转换方法
float_arr.dtype
dtype('float64')
浮点数转换为整型时,小数点会被省略
arr = np.array([3.7, -1.2, -2.6, 0.5, 12.9, 10.1])
arr
array([ 3.7, -1.2, -2.6, 0.5, 12.9, 10.1])
arr.astype(np.int32) #转换方法
array([ 3, -1, -2, 0, 12, 10], dtype=int32)
字符串转换为浮点数
n_s = np.array(['1.25','-9.6','42'],dtype=np.string_)
n_s.astype(float)#转换
array([ 1.25, -9.6 , 42. ])
2.3 直接替换
int_array = np.arange(10)
c_a = np.array([.22,.270,.357,.380,.44,.50],dtype=np.float64)
int_array.astype(c_a.dtype) #将c_a的数据类型应用到int_array
array([0., 1., 2., 3., 4., 5., 6., 7., 8., 9.])
运用简称可减少代码量
运算
arr = np.array([[1.,2.,3.,],[4.,5.,6.]])
arr
array([[1., 2., 3.],
[4., 5., 6.]])
arr * arr #自己*自己
array([[ 1., 4., 9.],
[16., 25., 36.]])
arr - arr
array([[0., 0., 0.],
[0., 0., 0.]])
1 / arr #标量
array([[1. , 0.5 , 0.33333333],
[0.25 , 0.2 , 0.16666667]])
arr * 0.5 #应用到每个元素
array([[0.5, 1. , 1.5],
[2. , 2.5, 3. ]])
arr2 = np.array([[0.,4.,1.],[7.,2.,12.]])
arr2 > arr #比较,必须在相同size
array([[False, True, False],
[ True, False, True]])
索引和切片
arr = np.arange(10)
arr
array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
arr[5] #选取,从0开始
5
arr[5:8] #切片
array([5, 6, 7])
arr[5:8] = 12 #赋值
arr
array([ 0, 1, 2, 3, 4, 12, 12, 12, 8, 9])
arr_slice = arr[5:8] #构建新对象
arr_slice
array([12, 12, 12])
arr_slice[1] = 12345 #新对象的变化会影响原来对象
arr
array([ 0, 1, 2, 3, 4, 12, 12345, 12, 8,
9])
arr_slice[:] = 64 #特殊选取方法
arr
array([ 0, 1, 2, 3, 4, 64, 64, 64, 8, 9])
arr[5:8].copy #显性表示复制
多维亦是如此.相应联想
arr2d = np.array([[1,2,3],[4,5,6],[7,8,9]])
arr2d[2]#索引
array([7, 8, 9])
arr2d[2][2]
9
arr2d[2,2] #同等方法
9
两维时,axis 0 代表 rows(行),axis 1 代表 columns(列)
arr3d = np.array([[[1,2,3],[4,5,6]],[[7,8,9],[10,11,12]]]) #2 2 3
arr3d
array([[[ 1, 2, 3],
[ 4, 5, 6]],
[[ 7, 8, 9],
[10, 11, 12]]])
arr3d[0] #依旧为 2 X 3的array对象
array([[1, 2, 3],
[4, 5, 6]])
old_values = arr3d[0].copy()
arr3d[0]=42
arr3d
array([[[42, 42, 42],
[42, 42, 42]],
[[ 7, 8, 9],
[10, 11, 12]]])
arr3d[0] = old_values #copy对象不影响原对象
arr3d
array([[[ 1, 2, 3],
[ 4, 5, 6]],
[[ 7, 8, 9],
[10, 11, 12]]])
arr3d[1,0,0] #嵌套选取
7
切片索引
arr
array([ 0, 1, 2, 3, 4, 64, 64, 64, 8, 9])
arr[1:6]
array([ 1, 2, 3, 4, 64])
arr[:6]
array([ 0, 1, 2, 3, 4, 64])
arr2d
array([[1, 2, 3],
[4, 5, 6],
[7, 8, 9]])
arr2d[:2] #止于2,不包括第2,选取前两行
array([[1, 2, 3],
[4, 5, 6]])
arr2d[:2,1:] #选取前两行的后两列
array([[2, 3],
[5, 6]])
arr2d[1,:2] #选取第一行的前两列
array([4, 5])
arr2d[:,:1] #冒号代表全部选中
array([[1],
[4],
[7]])
arr2d[:2,1:] = 0 #替换依然可行
arr2d
array([[1, 0, 0],
[4, 0, 0],
[7, 8, 9]])
布尔值,即真假,原理是真假对用1和0,只有两个,便于判断
names = np.array(['Bob','Joe','Will','Bob','Will','Joe','Joe'])
data = np.random.randn(7,4) #长度必须相等
names
array(['Bob', 'Joe', 'Will', 'Bob', 'Will', 'Joe', 'Joe'], dtype='data
array([[ 0.03317879, -0.64295019, -0.07074687, -0.47557693],
[-0.38937679, -1.12534115, -1.81676011, -1.15545174],
[-0.35255783, -0.20309579, -0.78609489, -0.83429794],
[-0.72824954, -0.01455318, 2.40890211, 0.45262685],
[-1.50436899, -0.01719232, -0.70454368, 2.29021736],
[-1.0090259 , -1.0858159 , -0.55730015, 0.29311532],
[ 0.24161904, 0.04861702, 0.23810033, -0.78593455]])
names == 'Bob' #两个等号表示数学意义上的相等,一个等号表示逻辑上的赋值
array([ True, False, False, True, False, False, False])
data[names == 'Bob']
array([[ 0.03317879, -0.64295019, -0.07074687, -0.47557693],
[-0.72824954, -0.01455318, 2.40890211, 0.45262685]])
data[names == 'Bob', 2:] #高级选取
array([[ 0.53825689, 0.08749286],
[-0.85205114, -0.772436 ]])
data[names == 'Bob',3]
array([ 0.08749286, -0.772436 ])
names != 'Bob' #取反
array([False, True, True, False, True, True, True])
data[~(names == 'Bob')] #取反
array([[-1.119334 , -0.99315582, -0.14021638, -0.68652541],
[ 0.66135045, -0.27269187, 0.25154645, 0.26779766],
[-0.42776762, -0.84351095, -0.79780832, 1.80460797],
[-1.1681968 , 0.13817726, 0.41940656, -0.15786619],
[ 0.42844477, -1.58866021, 0.31450085, -0.03915103]])
cond = names == 'Bob' #简化操作
data[~cond]
array([[-1.119334 , -0.99315582, -0.14021638, -0.68652541],
[ 0.66135045, -0.27269187, 0.25154645, 0.26779766],
[-0.42776762, -0.84351095, -0.79780832, 1.80460797],
[-1.1681968 , 0.13817726, 0.41940656, -0.15786619],
[ 0.42844477, -1.58866021, 0.31450085, -0.03915103]])
mask = (names == 'Bon') | (names == 'Will') #或
mask
array([False, False, True, False, True, False, False])
data[mask]
array([[ 0.66135045, -0.27269187, 0.25154645, 0.26779766],
[-0.42776762, -0.84351095, -0.79780832, 1.80460797]])
mask = (names == 'Bob') & (names == 'Will') #和
mask
array([False, False, False, False, False, False, False])
data[mask]
array([], shape=(0, 4), dtype=float64)
根据判断条件批量赋值
data[data < 0] = 0
data
array([[0. , 0. , 0.53825689, 0.08749286],
[0. , 0. , 0. , 0. ],
[0.66135045, 0. , 0.25154645, 0.26779766],
[1.34030771, 1.53001014, 0. , 0. ],
[0. , 0. , 0. , 1.80460797],
[0. , 0.13817726, 0.41940656, 0. ],
[0.42844477, 0. , 0.31450085, 0. ]])
data[names != 'Joe'] = 7
data
array([[7. , 7. , 7. , 7. ],
[0. , 0. , 0. , 0. ],
[7. , 7. , 7. , 7. ],
[7. , 7. , 7. , 7. ],
[7. , 7. , 7. , 7. ],
[0. , 0.13817726, 0.41940656, 0. ],
[0.42844477, 0. , 0.31450085, 0. ]])
花式索引
arr = np.empty((8,4))
for i in range(8):
arr[i] = i #按行赋值
arr
array([[0., 0., 0., 0.],
[1., 1., 1., 1.],
[2., 2., 2., 2.],
[3., 3., 3., 3.],
[4., 4., 4., 4.],
[5., 5., 5., 5.],
[6., 6., 6., 6.],
[7., 7., 7., 7.]])
arr[[4,3,0,6]] #直接选取行
array([[4., 4., 4., 4.],
[3., 3., 3., 3.],
[0., 0., 0., 0.],
[6., 6., 6., 6.]])
arr[[-3,-5,-7]] #倒序也可以
array([[5., 5., 5., 5.],
[3., 3., 3., 3.],
[1., 1., 1., 1.]])
arr = np.arange(32).reshape((8,4))
arr
array([[ 0, 1, 2, 3],
[ 4, 5, 6, 7],
[ 8, 9, 10, 11],
[12, 13, 14, 15],
[16, 17, 18, 19],
[20, 21, 22, 23],
[24, 25, 26, 27],
[28, 29, 30, 31]])
arr[[1,5,7,2],[0,3,1,2]] #行列对应选取,依次为第一行第0个,第五行第3个,,,,
array([ 4, 23, 29, 10])
见证奇迹的时刻,选取为矩阵
arr[[1,5,7,2]][:,[0,3,1,2]]
array([[ 4, 7, 5, 6],
[20, 23, 21, 22],
[28, 31, 29, 30],
[ 8, 11, 9, 10]])
行列转置
arr = np.arange(15).reshape((3,5))
arr
array([[ 0, 1, 2, 3, 4],
[ 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14]])
arr.T
array([[ 0, 5, 10],
[ 1, 6, 11],
[ 2, 7, 12],
[ 3, 8, 13],
[ 4, 9, 14]])
np.dot 内积
arr = np.random.randn(6,3)
arr
array([[ 0.14632913, 0.52363917, -0.61383088],
[-0.43408235, -0.23481596, 0.50205854],
[ 0.58277602, 0.39006049, 2.44427147],
[-1.60012924, 1.13015109, 1.87930724],
[ 0.06145772, -1.50099567, -0.21680612],
[-0.12640319, 1.1378038 , 0.33948385]])
np.dot(arr.T,arr)
array([[ 3.12963599, -1.63858658, -1.94666402],
[-1.63858658, 5.30631065, 3.34968866],
[-1.94666402, 3.34968866, 10.29736402]])
多维重排
arr = np.arange(16).reshape((2,2,4))
arr
array([[[ 0, 1, 2, 3],
[ 4, 5, 6, 7]],
[[ 8, 9, 10, 11],
[12, 13, 14, 15]]])
arr.transpose((1,0,2)) #重排方法
array([[[ 0, 1, 2, 3],
[ 8, 9, 10, 11]],
[[ 4, 5, 6, 7],
[12, 13, 14, 15]]])
arr
array([[[ 0, 1, 2, 3],
[ 4, 5, 6, 7]],
[[ 8, 9, 10, 11],
[12, 13, 14, 15]]])
arr.swapaxes(1,2) #另一方法
array([[[ 0, 4],
[ 1, 5],
[ 2, 6],
[ 3, 7]],
[[ 8, 12],
[ 9, 13],
[10, 14],
[11, 15]]])
ufunc 通用函数
arr = np.arange(10)
arr
array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
np.sqrt(arr) #平方根
array([0. , 1. , 1.41421356, 1.73205081, 2. ,
2.23606798, 2.44948974, 2.64575131, 2.82842712, 3. ])
np.exp(arr)
array([1.00000000e+00, 2.71828183e+00, 7.38905610e+00, 2.00855369e+01,
5.45981500e+01, 1.48413159e+02, 4.03428793e+02, 1.09663316e+03,
2.98095799e+03, 8.10308393e+03])
二元
x = np.random.randn(8)
y = np.random.randn(8)
x
array([ 0.76619326, -0.51913337, -0.61943584, -0.51367405, -1.81071209,
0.20368937, -0.20443369, -0.99255842])
y
array([ 0.69034544, 0.05340984, -0.96784714, -1.6425589 , -0.06741981,
0.71306132, -0.58463922, -1.3921061 ])
np.maximum(x,y) #取最大值,元素比较
array([ 0.76619326, 0.05340984, -0.61943584, -0.51367405, -0.06741981,
0.71306132, -0.20443369, -0.99255842])
arr = np.random.randn(7) * 5
arr
array([ 1.75351542, -6.57187357, 9.2276292 , 7.32215024, 5.6253457 ,
-5.31130902, -2.07810797])
r_m,w_p = np.modf(arr) #分离为整数部分和小数部分
r_m #小数部分
array([ 0.75351542, -0.57187357, 0.2276292 , 0.32215024, 0.6253457 ,
-0.31130902, -0.07810797])
w_p #整数部分
array([ 1., -6., 9., 7., 5., -5., -2.])
arr
array([ 1.75351542, -6.57187357, 9.2276292 , 7.32215024, 5.6253457 ,
-5.31130902, -2.07810797])
np.sqrt(arr) #负数不能开根
/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/ipykernel_launcher.py:1: RuntimeWarning: invalid value encountered in sqrt
"""Entry point for launching an IPython kernel.
array([1.32420369, nan, 3.0377013 , 2.7059472 , 2.37178112,
nan, nan])
np.sqrt(arr,arr)
/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/ipykernel_launcher.py:1: RuntimeWarning: invalid value encountered in sqrt
"""Entry point for launching an IPython kernel.
array([1.32420369, nan, 3.0377013 , 2.7059472 , 2.37178112,
nan, nan])
arr #新array对象
array([1.32420369, nan, 3.0377013 , 2.7059472 , 2.37178112,
nan, nan])
abs,fabs 绝对值
square 平方
log,log10,log2,log1p
isnan 判断
二元:add,subtract,multiply,divide, 加减乘除
maximum,minimum,
mod 取余
greater(_equal)/less 比较
logical_and/or/xor &/|/^
meshgrid
points = np.arange(-5,5,0.01)
xs,ys = np.meshgrid(points,points)
xs
array([[-5. , -4.99, -4.98, ..., 4.97, 4.98, 4.99],
[-5. , -4.99, -4.98, ..., 4.97, 4.98, 4.99],
[-5. , -4.99, -4.98, ..., 4.97, 4.98, 4.99],
...,
[-5. , -4.99, -4.98, ..., 4.97, 4.98, 4.99],
[-5. , -4.99, -4.98, ..., 4.97, 4.98, 4.99],
[-5. , -4.99, -4.98, ..., 4.97, 4.98, 4.99]])
ys
array([[-5. , -5. , -5. , ..., -5. , -5. , -5. ],
[-4.99, -4.99, -4.99, ..., -4.99, -4.99, -4.99],
[-4.98, -4.98, -4.98, ..., -4.98, -4.98, -4.98],
...,
[ 4.97, 4.97, 4.97, ..., 4.97, 4.97, 4.97],
[ 4.98, 4.98, 4.98, ..., 4.98, 4.98, 4.98],
[ 4.99, 4.99, 4.99, ..., 4.99, 4.99, 4.99]])
z = np.sqrt(xs ** 2 + ys ** 2)
z
array([[7.07106781, 7.06400028, 7.05693985, ..., 7.04988652, 7.05693985,
7.06400028],
[7.06400028, 7.05692568, 7.04985815, ..., 7.04279774, 7.04985815,
7.05692568],
[7.05693985, 7.04985815, 7.04278354, ..., 7.03571603, 7.04278354,
7.04985815],
...,
[7.04988652, 7.04279774, 7.03571603, ..., 7.0286414 , 7.03571603,
7.04279774],
[7.05693985, 7.04985815, 7.04278354, ..., 7.03571603, 7.04278354,
7.04985815],
[7.06400028, 7.05692568, 7.04985815, ..., 7.04279774, 7.04985815,
7.05692568]])
where 条件判断
xarr = np.array([1.1,1.2,1.3,1.4,1.5])
yarr = np.array([2.1,2.2,2.3,2.4,2.5])
cond = np.array([True,False,True,True,False])
result
[1.1, 2.2, 1.3, 1.4, 2.5]
result = np.where(cond,xarr,yarr) #谁真取谁
result
array([1.1, 2.2, 1.3, 1.4, 2.5])
正负值替换
arr = np.random.randn(4,4)
arr
array([[-0.14266064, -0.16213865, -0.58804574, 0.45092737],
[ 0.8518436 , -0.78783841, -0.62372951, 1.13642129],
[ 0.06346719, 1.20635278, -0.09238897, 1.73370433],
[ 1.62183372, 0.08720613, -0.63442454, 2.27619414]])
cond = arr > 0
np.where(cond, 2, -2) #符合条件为2,不符合为-2
array([[-2, -2, -2, 2],
[ 2, -2, -2, 2],
[ 2, 2, -2, 2],
[ 2, 2, -2, 2]])
np.where(cond,2,arr) #不符合保持原貌
array([[-0.14266064, -0.16213865, -0.58804574, 2. ],
[ 2. , -0.78783841, -0.62372951, 2. ],
[ 2. , 2. , -0.09238897, 2. ],
[ 2. , 2. , -0.63442454, 2. ]])
数学和统计方法
arr = np.random.randn(5,4)
arr
array([[-0.68218677, 1.10310301, -0.21057012, -0.55508466],
[ 0.91256439, -0.20883975, -0.09176431, -0.66631307],
[-1.04120502, -1.11187463, -0.03471294, -0.11266913],
[ 1.71228664, 0.23509329, 0.31352707, -2.61562304],
[-0.91986032, 0.5552931 , 0.48340178, -0.20594334]])
arr.mean() #平均值
-0.15706889233659155
np.mean(arr) #平均值
-0.15706889233659155
arr.sum() #和
-3.1413778467318307
行列统计
arr.mean(axis=1) #按列取均值
array([-0.08618463, -0.01358819, -0.57511543, -0.08867901, -0.0217772 ])
arr.sum(axis=0) #按行取和
array([-0.01840108, 0.57277501, 0.45988147, -4.15563325])
其他方法
arr = np.array([0,1,2,3,4,5,6,7])
arr
array([0, 1, 2, 3, 4, 5, 6, 7])
arr.cumsum() #累计和
array([ 0, 1, 3, 6, 10, 15, 21, 28])
多维
arr = np.array([[0,1,2],[3,4,5],[6,7,8]])
arr
array([[0, 1, 2],
[3, 4, 5],
[6, 7, 8]])
arr.cumsum(0) #按行取累计和
array([[ 0, 1, 2],
[ 3, 5, 7],
[ 9, 12, 15]])
arr.cumprod(1) #按列
array([[ 0, 0, 0],
[ 3, 12, 60],
[ 6, 42, 336]])
std 标准差
var 方差
min,max 大小
argmin,argmax 返回索引
布尔值方法
arr = np.random.randn(100)
(arr > 0).sum() #判断条件
63
bools = np.array([False,False,True,False])
bools.any() #一真则真,全假则假
True
bools.all() #全真为真,一假则假
False
排序
arr = np.random.randn(6)
arr
array([-1.19241559, 1.86955897, -1.47747941, -0.49292942, 1.38922409,
-0.07618703])
arr.sort() #排序
arr
array([-1.47747941, -1.19241559, -0.49292942, -0.07618703, 1.38922409,
1.86955897])
多维行列排序
arr = np.random.randn(5,3)
arr
array([[-0.18284374, -1.92183403, -1.33962176],
[-0.35250898, 0.83480866, 0.43398843],
[ 0.04709364, -0.08379553, -0.09993549],
[ 0.87922406, 1.75488263, -0.57784417],
[-0.93287139, -0.11406265, 0.71589254]])
arr.sort(1) #列
arr
array([[-1.92183403, -1.33962176, -0.18284374],
[-0.35250898, 0.43398843, 0.83480866],
[-0.09993549, -0.08379553, 0.04709364],
[-0.57784417, 0.87922406, 1.75488263],
[-0.93287139, -0.11406265, 0.71589254]])
特定位置
l_a = np.random.randn(1000)
l_a.sort()
l_a[int(0.05 * len(l_a))] #5%
-1.7231654793957891
独一无二
names
array(['Bob', 'Joe', 'Will', 'Bob', 'Will', 'Joe', 'Joe'], dtype='np.unique(names)
array(['Bob', 'Joe', 'Will'], dtype='ints = np.array([3,3,3,2,2,1,1,4,4])
np.unique(ints)
array([1, 2, 3, 4])
pure python
sorted(set(names))
检验
values = np.array([6,0,0,3,2,5,6])
np.in1d(values,[2,3,6]) #检验values中的值是否在后面的list中
array([ True, False, False, True, True, False, True])
文件输入与输出
arr = np.arange(10)
np.save('/Users/meininghang/Desktop/s_a.npy',arr) #存储文件
np.load('/Users/meininghang/Desktop/s_a.npy') #读取
array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
np.savez('/Users/meininghang/Desktop/s_a.npz',a=arr,b=arr) #存储文件为.npz
arch = np.load('/Users/meininghang/Desktop/s_a.npz') #读取文件
arch['b']
array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
arch['a']
array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
np.savez_compressed('/Users/meininghang/Desktop/s_a2.npz',a = arr,b=arr) #存储为压缩文件
线代
x = np.array([[1,2,3],[4,5,6]])
y = np.array([[6,23,],[-1,7],[8,9]])
x
array([[1, 2, 3],
[4, 5, 6]])
y
array([[ 6, 23],
[-1, 7],
[ 8, 9]])
x.dot(y)
array([[ 28, 64],
[ 67, 181]])
np.dot(x,y) #等价
array([[ 28, 64],
[ 67, 181]])
x @ np.ones(3) #特殊用法
array([ 6., 15.])
等价于np.dot(x,np.ones(3))
np.ones(3)
array([1., 1., 1.])
其他
from numpy.linalg import inv,qr
X = np.random.randn(5,5)
mat = X.T.dot(X)
inv(mat) #取反
array([[ 0.3153804 , 0.06572043, 0.09295758, -0.05295502, 0.10039062],
[ 0.06572043, 0.14119364, 0.0532526 , -0.04233026, 0.02636569],
[ 0.09295758, 0.0532526 , 0.41001573, 0.00487669, 0.00854835],
[-0.05295502, -0.04233026, 0.00487669, 0.13017746, -0.01152687],
[ 0.10039062, 0.02636569, 0.00854835, -0.01152687, 0.1324641 ]])
mat.dot(inv(mat))
array([[ 1.00000000e+00, 1.38314235e-18, -2.30089733e-17,
-2.31625676e-18, 0.00000000e+00],
[-4.36359852e-17, 1.00000000e+00, 1.24809158e-17,
-6.64902607e-17, -2.77555756e-17],
[ 4.67942395e-18, 1.06484372e-17, 1.00000000e+00,
5.01064367e-19, -1.38777878e-17],
[ 6.45223305e-17, -1.90564222e-17, 1.49753705e-17,
1.00000000e+00, 2.77555756e-17],
[ 0.00000000e+00, -5.55111512e-17, 1.38777878e-17,
1.38777878e-17, 1.00000000e+00]])
q,r = qr(mat)
q
array([[-0.78454422, 0.03130395, -0.09538485, 0.14529988, 0.59439053],
[ 0.13807729, -0.9415313 , -0.09623544, 0.24662385, 0.15610539],
[ 0.15150193, 0.11585421, -0.97871075, -0.05646742, 0.05061286],
[-0.23236293, -0.28614052, -0.01989414, -0.92686434, -0.06824804],
[ 0.53710302, 0.13130224, 0.15287236, -0.23621718, 0.78429048]])
r
array([[-6.32722241, 0.7479685 , 1.57470985, -3.43964677, 8.30010971],
[ 0. , -9.16278613, 1.48489447, -5.03119444, 2.28135854],
[ 0. , 0. , -2.41505909, 0.05405567, 1.31462198],
[ 0. , 0. , 0. , -7.33442291, -2.42148731],
[ 0. , 0. , 0. , 0. , 5.92077767]])
random
samples = np.random.normal(size=(4,4))
samples
array([[ 0.64867627, 0.53180508, -0.54734626, -0.45303299],
[-0.03251697, 0.39046217, 1.98828937, 0.57414828],
[-2.342141 , -0.16672602, -0.08988844, -1.1986827 ],
[-1.52778635, 0.50663291, -1.06283763, -1.43748063]])
np.random.seed(1234)
rng = np.random.RandomState(1234) #isolate
rng.randn(10)
array([ 0.47143516, -1.19097569, 1.43270697, -0.3126519 , -0.72058873,
0.88716294, 0.85958841, -0.6365235 , 0.01569637, -2.24268495])
随机漫步
pure Python
import random
position = 0
walk = [position]
steps = 1000
for i in range(steps):
step = 1 if random.randint(0,1) else -1
position += step
walk.append(position)
import matplotlib.pyplot as plt
plt.plot(walk[:100])
[]
[图片上传失败...(image-2e107-1517836465428)]
nsteps = 1000
draws = np.random.randint(0,2,size = nsteps)
steps = np.where(draws>0,1,-1)
walk = steps.cumsum()
walk.min()
-9
walk.max()
60
(np.abs(walk) >= 10).argmax()
297