There are two ways to create an array, 1. Direct assignment 2. Generation of random variables
random generation includes 4 types: Np.arange, Np.linspace (0,10,5), Np.logspace (0,2,5), Np.random.random (3,2,3)
Np.arange (10,20,2) # #左闭右开区间, start value, end value, step
Np.linspace (0,10,5) # #闭区间, starting value, terminating value, number of elements arithmetic progression
Np.logspace (0,2,5) # #闭区间, starting value (in exponential form), terminating value (in exponential form, base 10, 2 exponent), number of elements geometric series
np.random.random (3,2,3) #三维是3个, two-dimensional is 2, one-dimensional is 3
properties of the Ndarray:
The DYTPE output is the data type of the element that makes up the array, int+32
shape a tuple of each dimension size of an array, such as return (2, 5)
total number of size elements
Ndim The number of dimensions of an array, such as a three-dimensional array return is 3
Ndarry Modify the shape (only shape will change the original data, others will not):
A.reshape ( -1,5) #这里-1 means adaptive, 5 means 5 columns
A.shape (2,-1) #方法中的值同reshape, but shape changed the original array, reshape did not change the original
modifying the values in an array can be sliced
through the transpose transformation array, such as array shape from (5,8) can be converted to shape (8,5), just extract the data, the original data unchanged
convert Direct shape from (5,8) to (8,5) by property T, just extract data, original data unchanged
Subtraction between arrays, arrays and arrays , between arrays and scalars (numbers)
The matrix product of an array , with one column of values multiplied by one column of another array
Array Index of a Boolean type
Fancy Index Arr[np.ix_ ([0,3,5],[0,2,3])
general functions, one-dimensional functions, two-dimensional functions
aggregate function, Min min, max Max, mean average, etc., such as arr.min (), return a specific value
In a two-dimensional array, axis=0 represents a column, Axis=1 represents a row, such as Arr.min (axis=0), an aggregation of elements on the same column, and a row that returns the smallest value of the column.
where you can find replacements for the specified element np.where (condition, replace set value, array)
unique can exclude duplicates and return a one-dimensional array
Please forgive me, this writing is really not for people to see ... Example later look mood add
Python simple data cleansing, data filtering method collation