NumPy Basics: Arrays and vector calculations

The boss fire today, or continue to transcription it, to appease my wounded little heart. Knowledge still has to accumulate slowly, step by step, this perhaps is the quickest shortcut.

------2015-2-16------------------------------------------------------------------

• NumPy Ndarray: A multidimensional Array object

An important feature of NumPy is the n-dimensional array object (Ndarray), which is a fast and flexible large data set container. Ndarray is a generic homogeneous data multidimensional container, which means that all elements must be of the same type. Each array has a shape (a tuple that represents the size of each dimension) and Dtype (an object representing the array data type).

Import NumPy as npin[3]: data=[[1,2,3],[4,5,6]]in[4]: arr=np.array (data) in[6]: arrout[ 6]: Array ([[1, 2, 3],       [4, 5, 6]]) in[7]: arr.shapeout[7]: (2L, 3L) in[ 8]: arr.dtypeout[8]: Dtype ('int32')

Create Ndarray

Array creation function

Function	Description
Array	Converts the input data (list, tuple, array, or other sequence type) to Ndarray. Either infer the Dtype, or display the specified dtype. Default direct copy of input data
Asarray	Converts the input to Ndarray if the input itself is a ndarray and does not replicate
Arange	Returns a ndarray instead of a list
Ones, Ones_like	Creates a full 1 array based on the specified shape and dtype. Ones_like takes another array as an argument and creates a full 1 array based on its shape and dtype
Zeros, Zeros_like	Similar to ones and ones_likes only produce a full 0 array
Empty, Empty_like	Creates a new array, allocates only memory space but does not populate any values
Eye, identity	Create a NXN unit matrix

Ndarray Data types

Int8,int16,int32,int64 signed integral type

Uint8,uint16,uint32,uint64 unsigned integral type

float16,float32,float64,float128 single precision, multi-precision, extended accuracy

The plural of complex64,complex128,complex256, respectively, with 32,64,128

bool

Object Python Data Objects

String_ fixed-length string data types

Unicode_ fixed-length Unicode types

In[23]: Arr.astype (np.float64) out[: Array ([1.,  2.,  3.,  4.,  5.]) in[]: h1=arr.astype (np.int16) in[[]: h1.dtypeout[[]: Dtype ('int16 ')

Operations between arrays and scalars

IN[2]:ImportNumPy as npin[3]: Arr=np.array ([[1,2,3],[4,5,6]]) in[4]: arr*arrout[4]: Array ([[1, 4, 9],       [16, 25, 36]]) in[5]: arr+arrout[5]: Array ([[2, 4, 6],       [ 8, 10, 12]]) in[6]: arr*4out[6]: Array ([[4, 8, 12],       [16, 20, 24]]) in[7]: arr**0.5out[7]: Array ([[1., 1.41421356, 1.73205081],       [ 2., 2.23606798, 2.44948974]])

Basic indexes and slices

IN[8]: Np.arange (ten) out[8]: Array ([0, 1, 2, 3, 4, 5, 6, 7, 8, 9]) in[9]: Arr=np.arange (ten) in[10] : Arr[5:8]out[]: Array ([5, 6, 7]) in[]: arr_slice=arr[5:8]in[]: arr_slice[1]out[ ]: 6in[]: arr_slice[1]=123456in[]:arrout[]: Array ([     0,      1,      2,      3,      4,      5, 123456,      7,            8,      9])

Warning: Ndarray a copy of a slice rather than a view, you need to display arr[5:8].copy ()

Transpose of arrays and axes swapping

IN[16]: Arr=np.arange (0). Reshape ((3,5)) in[[]: arrout[]: Array (  [[[1], "  2", "["]] "." [[[]]].  3,  4],       5,  6,  7,  8,  9],       [Ten, one, one, and up]]) in[ ]: arr. tout[]: Array ([[0  ,5, ten],       1,  6, one),       2,  7, 12 ],       3,  8, [],       4,  9,]]) in[]: Np.dot (Arr,arr. T) out[         , 80, 255, 430],

• General functions: Fast element Progression Group functions

• Using Arrays for data processing

• File input and output for arrays

• Linear algebra

• Random number generation

IN[20]: Samples=np.random.normal (size= (bis)) in[]: samplesout[1.2160082,  0.34629744, -0.70813727,  2.59673398],       [-1.32110632,  1.19660352,  0.08227731,  0.24075048],       [-0.29301216,  0.42639032, -1.76321448, -1.05558718],       0.0872803 ,  0.25871173,  0.63373105,  0.59362002])

The Numpy.random module is faster than the python built-in random module.

Partial numpy.random function

NumPy Basics: Arrays and vector calculations