Python Cheat Sheet for NumPy

Built-in Functions
S.No Function Name Example Output
1 array() a1= np.array([1,2,3,4])
a2=np.array([[2,4,6],[1,3,5]])
a1.shape (4,)
a2.shape (2,3)
a2.itemsize 4
list1=[1,2,3,4,5]
ar1=np.array(list1)
print(type(ar1)) <class ‘numpy.ndarray’>
print(ar1.dtype) Int32
print(ar1.itemsize) 4
print(ar1) [1 2 3 4 5]
print(ar1[0],ar1[3]) 14
tup1=(2.5,3.5,4.5,6.5)
ar2=np.array(tup1)
print(type(ar2)) <class ‘numpy.ndarray’>
print(ar2.dtype) float64
print(ar2.itemsize) 8
print(ar2) [2.5 3.5 4.5 6.5]
print(ar2[0],ar2[3]) 2.5 6.5
ar3=np.array(list1,dtype=int64)
print(ar3.dtype) int64
print(ar3.itemsize) 8
ar4=np.array(tup1,dtype=np.float16)
 print(ar4.dtype) float16
print(ar4.itemsize) 2
2 fromiter() adict={1:”A”,2:”B”,3:”C”,4:”D”,5:”E”}
ar5=np.fromiter(adict,dtype=np.int32)
ar5.dtype,ar5.itemsize dtype(‘int32’),4)
print(ar5[0],ar5[3]) 1 4
print(ar5) [1 2 3 4 5]
alist=[‘a’,’b’,’c’,’d’,’e’]
atup=(1.5,2.5,3.5,4.5)
lst2=[a*2-3 for a in range(2,6)]
nd1=np.fromiter(alist,dtype=”U1”) array([‘a’,’b’,’c’,’d’,’e’],dtype=”<U1”)
nd2=np.fromiter(atup,dtype=”U1”) array([[‘1’,’2’,’3’,’4’],dtype=”<U1”)
nd3=np.fromiter(lst2,dtype=np.int16) array([1,3,5,7],dtype=int16)
3 arange() arr5=np.arange(7) array([0,1,2,3,4,5,6])
arr6=np.arange(1,7,2,np.float32) array([1.,3.,5.],dtype=float32)
4 linspace() arr1=np.linspace(2.5,5,6) array([2.5,3.,3.5,4.,4.5,5.])
5 reshape() ar=np.arange(10) array([0,1,2,3,4,5,6,7,8,9])
ar1=np.reshape(5,2) array([[0,1],
[2,3],
[4,5],
[6,7],
[8,9]])
ar2=ar.reshape(2,5) array([0,1,2,3,4],
[5,6,7,8,9]])
ary=np.arange(8.0).reshape(2,4) array([[0.,1.,2.,3.],
[4., 5.,6.,7.]])
6 empty() arr1=np.empty([3,2]) If you display arr1,arr2, it will display any random
arr2=np.empty([3,4],dtype=np.int8) contents, which are uninitialized garbage values.
7 zeros() arr3=np.zeros([2,3],dtype=np.int64,order=”c”) array([0,0,0],
[0,0,0]],dtype=int64)
8 ones() arr4=np.ones([2,3],dtype=np.float32) array([1.,1.,1.],
[1.,1.,1.]],dtype=float32)
9 empty_like() arr5=np.empty_like(arr4) If you display ar5 it will display any random contents,
which are uninitialized garbage values.
10 zeros_like() arr6=np.zeros_like(arr1) array([[0,0],
[0,0],
[0,0]])
11 ones_like() arr7=np.ones_like(arr3) array([[1,1,1],
[1,1,1]],dtype=int64)
12 eye() np.eye(5) Create a 5 x 5 array with values 0 at non-diagonal
positions and with 1 on diagonal
13 full() np.full((2,3),6) Create a 2 x 3 array with all values as 6
14 rand() np.random.rand(3,4) Create a 3 x 4 array of random floats in the range 0-1
np.random.rand(4,5)*100 Create a 4 x 5 array of random floats in the rage 0-100
15 randint() np.random.randint(5,size=(2,3)) Create a 2 x 3 array with random ints in the range 0-4
16 vstack() Arr1=np.array([[0,1,2],
[3,4,5],
[6,7,8]])
Arr2=np.array([[10,11,12],
[13,14,15],
[16,17,18]])
Arr3=np.vstack((Arr1,Arr2)) array([[0,1,2],
[3,4,5],
[6,7,8],
[10,11,12],
[13,14,15],
 [16,17,18]])
17 Hstack() Arr4=np.hstack((Arr1,Arr2)) array([[0,1,2,10,11,12],
[3,4,5,13,14,15],
[6,7,8,16,17,18]])
18 concatenate()
jar1=np.concatenate((ar1,ar2),axis=0) array([[3,4,1],

ar1= [6,8,5],
[2,1,3],
[2,4,2],
3 4 1
[1,9,1]])
6 8 5
2 1 3

jar2=np.concatenate((ar1,ar3),axis=1) array([[3,4,1,6,1],

ar2= [6,8,5,2,1],
[2,1,3,8,5]])
2 4 2
1 9 1 jar3=np.concatenate((ar1,ar4),axis=1) array([[3,4,1,2],
[6,8,5,8],

ar3=
[2,1,3,5]])

jar4=np.concatenate((ar2,ar1),axis=0) array([[2,4,2],
6 1
[1,9,1],
2 1
[3,4,1],
8 5
[6,8,5],
[2,1,3]])

ar4=
 2 jar5=np.concatenate((ar2,ar5),axis=1) array([[2,4,2,1,2,8,9],
8 [1,9,1,6,3,4,5]])
5 array([[3,4,1,2,1],
[6,8,5,4,9],
jar6=np.concatenate((ar1,ar2.T),axis=1) [2,1,3,2,1]])

ar5= array([[3,4,1,6,8,5,2,1,3,2,8,5])

1 2 8 9 jar7=np.concatenate((ar1,ar4),axis=None)
6 3 4 5

19 hsplit() ary

np.hsplit(ary,2)

0. 1. 2. 3. 4. 5.
6. 7. 8. 9. 10 11.
array([[0.,1.,2.],
12. 13. 14. 15. 16. 17. [6.,7.,8.],
18. 19. 20. 21. 22. 23. [12.,13.14.],
[18.,19.,20.]])
 array([[3.,4.,5.],
[9.,10.,11.],
[15.,16.,17.],
[21.,22.,23.]])

np.hsplit(ary,3) array([[0.,1.],
[6.,7.],
[12.,13.],
[18.,19.]])

array([[2.,3.],
[8.,9.],
[14.,15.],
[20.,21.]])

array([[4.,5.],
[10.,11.],
[16.,17.],
[22.,23.]])

20 vsplit() ar1,ar2=np.vsplit(ary,2) >>>ar1

array([[0.,1.,2.,3.,4.,5.],
[6.,7.,8.,9.,10.,11.]])
>>>ar2
array([[12.,13.,14.,15.,16.,17.],
[18.,19.,20.,21.,22.,23.]])
 arid=[10,11,12,13,14,15,16,17,18,19] array([10,11]),array([12,13,14,15]),array([16,17,18,19])
21 split() np.split(arid,[2,6])
[2,6] can be understood as [0:2],[2:6],[6:]

ary

0. 1. 2. 3. 4. 5.
6. 7. 8. 9. 10 11.

12. 13. 14. 15. 16. 17.

18. 19. 20. 21. 22. 23.
np.split(ary,[1,4]) array([[0.,1.,2.,3.,4.,5.]])
can be understood as [0:1],[1:4],[4:]
array([[6.,7.,8.,9.,10.,11.],
[12.,13.,14.,15.,16.,17.,],
[18.,19.,20.,21.,22.,23.]])

22 extract() cond1=