numpy 查找 返回索引_Numpy 技巧总结

快速入门教程 | NumPy 中文​www.numpy.org.cn

创建保存

np.genfromtxt(data, ...) 链接

np.genfromtxt(data, ...)
>>> data = BytesIO("So it goesn#a b cn1 2 3n 4 5 6")
>>> np.genfromtxt(data, skip_header=1, names=True)
array([(1.0, 2.0, 3.0), (4.0, 5.0, 6.0)],
      dtype=[('a', '<f8'), ('b', '<f8'), ('c', '<f8')])

>>> data = "N/A, 2, 3n4, ,???"
>>> kwargs = dict(delimiter=",",
...               dtype=int,
...               names="a,b,c",
...               missing_values={
    0:"N/A", 'b':" ", 2:"???"},
...               filling_values={
    0:0, 'b':0, 2:-999})
>>> np.genfromtxt(BytesIO(data), **kwargs)
array([(0, 2, 3), (4, 0, -999)],
      dtype=[('a', '<i8'), ('b', '<i8'), ('c', '<i8')])

load(file[, mmap_mode, allow_pickle, …])
    Load arrays or pickled objects from .npy, .npz or pickled files.

save(file, arr[, allow_pickle, fix_imports])
    Save an array to a binary file in NumPy .npy format.

savez(file, *args, **kwds)
    Save several arrays into a single file in uncompressed .npz format.

savez_compressed(file, *args, **kwds)
    Save several arrays into a single file in compressed .npz format.

ndarray.tofile(fid[, sep, format])
    Write array to a file as text or binary (default).

ndarray.tolist()
    Return the array as an a.ndim-levels deep nested list of Python scalars.

索引查找

k维的数组，索引时需要k个索引值，索引值可以是常数、切片、数组索引（整数、布尔）

a = array([[ 0,  1,  2,  3],
           [ 4,  5,  6,  7],
           [ 8,  9, 10, 11]])
# a.shape = (3, 4)

# 整数及切片索引
# Slice用的是view的方式，而index用的是copy方式
a[1,2]    # 6
a[0, [1,2]]    # [1, 2]
a[0:2, 2:4]    # [[2, 3], [6, 7]]
a[:, 2:4]      # [[2, 3], [6, 7], [10, 11]]
a[0:1, ..., 2:4, 0] 

# 整数数组索引
i=array([[1, 2], [0, 1]])     
j=array([[0, 2], [1, 2]])      
a[i, j] == array([[ 2,  5],
                 [ 7, 11]])
# a[i, j].shape == a[0, j].shape == a[i, 0].shape == (2, 2)  

a[:, j] == array([[[ 2,  1],
                  [ 3,  3]],
                 [[ 6,  5],
                  [ 7,  7]],
                 [[10,  9],
                  [11, 11]]])
# a[:, j].shape == (3, 2, 2)
a[i, :] == array([[[ 0,  1,  2,  3],
                   [ 4,  5,  6,  7]],
                  [[ 4,  5,  6,  7],
                   [ 8,  9, 10, 11]]])
# a[i, :].shape == (2, 2, 4)

# 布尔数组索引
b = a > 4
b == array([[False, False, False, False],
            [False,  True,  True,  True],
            [ True,  True,  True,  True]], dtype=bool)
a[b] == array([ 5,  6,  7,  8,  9, 10, 11])
a[b] = 0  # 只是视图，可以赋值
a == array([[0, 1, 2, 3],
            [4, 0, 0, 0],
            [0, 0, 0, 0]])
a[a[:,0]>2, a[0,:]>1] # a中第一列大于2的行且第一行大于1的列的元素

# ix_函数可以产生多元组合
a = array([2,3,4,5])
b = array([8,5,4])
c = array([5,4,6,8,3])
ax,bx,cx = ix_(a,b,c)
# ax.shape, bx.shape, cx.shape == ((4, 1, 1), (1, 3, 1), (1, 1, 5))

result = ax+bx*cx  #利用广播机制，产生不同维度的所有组合
result[3,2,4] == a[3]+b[2]*c[4]  == 17

# 笛卡尔积
numpy.transpose([numpy.tile(x, len(y)), numpy.repeat(y, len(x))])
array([[1, 4],
       [2, 4],
       [3, 4],
       [1, 5],
       [2, 5],
       [3, 5]])

[[x0, y0] for x0 in x for y0 in y]

# 查找元素
np.where(condition, [x, y])  #[x,y]为可选项，满足条件返回x，不满足条件返回y
>>> np.where(a < 4, a, -1)  # -1 is broadcast
array([[ 0,  1,  2],
       [ 0,  2, -1],
       [ 0,  3, -1]])

np.argwhere()  #返回满足括号内条件的元素的索引，数组形式
>>> np.argwhere(a>5)
array([[0, 2],
       [1, 0],
       [1, 3],
       [2, 3]], dtype=int64)

数组遍历

# 获取索引迭代器
>>> for index in np.ndindex(3, 2, 1):
 |  ...     print(index)
 |  (0, 0, 0)
 |  (0, 1, 0)
 |  (1, 0, 0)
 |  (1, 1, 0