ML-using-Python-notes

Day-1(06/08/25): #1 importing required modules !pip install numpy import numpy as np #create nd-array from a list data1 =[11, 22,33,44.0,55,66,77,88,'AEC',True,False] print(data1,len(data1),type(data1)) arr1 =np.array(data1)

print(arr1, len(arr1),type(arr1), arr1.dtype) o/p-Successfully installed numpy-2.3.2 [11, 22, 33, 44.0, 55, 66, 77, 88, 'AEC', True, False] 11 <class 'list'> ['11' '22' '33' '44.0' '55' '66' '77' '88' 'AEC' 'True' 'False'] 11 <class 'numpy.ndarray'> <U32 #2 data1 = range(1,20,3) #list containing 0 to 9 print(data1) arr1 = np.array(data1) print(arr1) o/p-range(1, 20, 3) [ 1 4 7 10 13 16 19] #3

data2 =[range(1,5),range(6,10)] print(data2) arr2 = np.array(data2) print(arr2)#creating 2-D array for item in np.nditer(arr2): print(item, end=",") o/p-[range(1, 5), range(6, 10)] [[1 2 3 4] [6 7 8 9]] 1,2,3,4,6,7,8,9,

#4 data1 = [11,22,33,44] arr1=np.array(data1) print(data1,arr1) result=list(arr1) print(arr1,result) result= arr1.tolist() print(arr1,result) o/p [11, 22, 33, 44] [11 22 33 44] [11 22 33 44] [np.int64(11), np.int64(22), np.int64(33), np.int64(44)] [11 22 33 44] [11, 22, 33, 44] #5 data2 =[[11,22,33,44],[55,66,77,88]] arr2 =np.array(data2) print(data2) print(arr2) result =list(arr2) print(arr2) print(result) result= arr2.tolist() print(arr2) print(result) o/p-[[11, 22, 33, 44], [55, 66, 77, 88]] [[11 22 33 44] [55 66 77 88]] [[11 22 33 44] [55 66 77 88]] [array([11, 22, 33, 44]), array([55, 66, 77, 88])] [[11 22 33 44] [55 66 77 88]] [[11, 22, 33, 44], [55, 66, 77, 88]] #6 #examine arrays arr1 = np.array([11,22,33,44,55,66,77,88]) print(arr1, type(arr1),id(arr1)) print(arr1.dtype, arr1.shape,len(arr1),arr1.ndim) print(arr1.dtype,arr1.shape,len(arr1),arr1.ndim) print(arr1.size) o/p-[11 22 33 44 55 66 77 88] <class 'numpy.ndarray'> 1808502471536 int64 (8,) 8 1 int64 (8,) 8 1 8 #7 var1=(8) print(var1, type(var1)) var1 =(8,)#singleton representation of a tuple print(var1,type(var1)) o/p=8 <class 'int'> (8,) <class 'tuple'> #8 arr2 = np.array([[11,22,33,44],[55,66,77,88]]) print(arr2,type(arr2),id(arr2)) print(arr2.size) o/p-[[11 22 33 44] [55 66 77 88]] <class 'numpy.ndarray'> 1808502467888 8 #9 #create special arrays print(np.zeros(10)) print(np.zeros(10,np.int32)) print(np.zeros(10).astype(int)) print(np.zeros((2,5))) print(np.zeros((2,5),np.int32)) print(np.zeros((2,5)).astype(int)) o/p-[0. 0. 0. 0. 0. 0. 0. 0. 0. 0.] [0 0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 0] [[0. 0. 0. 0. 0.] [0. 0. 0. 0. 0.]] [[0 0 0 0 0] [0 0 0 0 0]] [[0 0 0 0 0] [0 0 0 0 0]] #10 print(np.zeros((3,5))+5) #vector and scalar operation print(np.ones((3,5)) *5) o/p-[[5. 5. 5. 5. 5.] [5. 5. 5. 5. 5.] [5. 5. 5. 5. 5.]] [[5. 5. 5. 5. 5.] [5. 5. 5. 5. 5.] [5. 5. 5. 5. 5.]] #11 print(np.linspace(1,10,5)) print(np.linspace(1,50,11)) o/p-[ 1. 3.25 5.5 7.75 10. ] [ 1. 5.9 10.8 15.7 20.6 25.5 30.4 35.3 40.2 45.1 50. ] #12 print(np.logspace(0,3,4)) print(np.logspace(0,3,4,base=2)) print(np.logspace(0,8,9,base = np.e)) o/p-[ 1. 10. 100. 1000.] [1. 2. 4. 8.] [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] #13 arr1 = np.arange(1,11,2) print(arr1,type(arr1)) arr1 = np.arange(11,0,-2) print(arr1,type(arr1)) arr1 = np.arange(1,11,2).astype(float) print(arr1,type(arr1)) o/p-[1 3 5 7 9] <class 'numpy.ndarray'> [11 9 7 5 3 1] <class 'numpy.ndarray'> [1. 3. 5. 7. 9.] <class 'numpy.ndarray'> #14 #reshape of ndarray print(np.arange(20)) print(np.arange(20).reshape(4,5)) print(np.arange(20).reshape(4,-1)) print(np.arange(20).reshape(-1,5)) o/p-[ 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19] [[ 0 1 2 3 4] [ 5 6 7 8 9] [10 11 12 13 14] [15 16 17 18 19]] [[ 0 1 2 3 4] [ 5 6 7 8 9] [10 11 12 13 14] [15 16 17 18 19]] [[ 0 1 2 3 4] [ 5 6 7 8 9] [10 11 12 13 14] [15 16 17 18 19]] #15 print(np.arange(20).reshape(2,-1,5)) o/p-[[[ 0 1 2 3 4] [ 5 6 7 8 9]]

[[10 11 12 13 14] [15 16 17 18 19]]] #16 arr2=np.arange(20).reshape(2,10) print(arr2) arr1=arr2.flatten() print(arr1) o/p- [[ 0 1 2 3 4 5 6 7 8 9] [10 11 12 13 14 15 16 17 18 19]] [ 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19] #17 arr2 = np.arange(20).reshape(5,4) print(arr2) arr_transpose= arr2.T print(arr_transpose) arr_transpose = arr2.transpose() print(arr_transpose) o/p-[[ 0 1 2 3] [ 4 5 6 7] [ 8 9 10 11] [12 13 14 15] [16 17 18 19]] [[ 0 4 8 12 16] [ 1 5 9 13 17] [ 2 6 10 14 18] [ 3 7 11 15 19]] [[ 0 4 8 12 16] [ 1 5 9 13 17] [ 2 6 10 14 18] [ 3 7 11 15 19]] #18 print(np.linspace(1,10,5)) print(np.linspace(1,50,11)) o/p-[ 1. 3.25 5.5 7.75 10. ] [ 1. 5.9 10.8 15.7 20.6 25.5 30.4 35.3 40.2 45.1 50. ] #19 arr2 = np.arange(20).reshape(5, 4) print(arr2) arr_transpose = arr2.T print(arr_transpose) arr_transpose = arr2.transpose() print(arr_transpose) o/p-[[ 0 1 2 3] [ 4 5 6 7] [ 8 9 10 11] [12 13 14 15] [16 17 18 19]] [[ 0 4 8 12 16] [ 1 5 9 13 17] [ 2 6 10 14 18] [ 3 7 11 15 19]] [[ 0 4 8 12 16] [ 1 5 9 13 17] [ 2 6 10 14 18] [ 3 7 11 15 19]] #20 matrix = np.arange(10,dtype=float).reshape((2,5)) print(matrix) print("stored in the C-tyle order(Row major):") cmatrix = matrix.copy(order='C') print(cmatrix) for item in np.nditer(cmatrix): print(item, end=",")
o/p-[[0. 1. 2. 3. 4.] [5. 6. 7. 8. 9.]] stored in the C-tyle order(Row major): [[0. 1. 2. 3. 4.] [5. 6. 7. 8. 9.]] 0.0,1.0,2.0,3.0,4.0,5.0,6.0,7.0,8.0,9.0, #21 matrix = np.arange(10,dtype=float).reshape((2,5)) print(matrix) print("stored in the F-style order(Column major):") fmatrix = matrix.copy(order='F') print(fmatrix) for item in np.nditer(fmatrix): print(item, end=",")
[[0. 1. 2. 3. 4.] [5. 6. 7. 8. 9.]] stored in the F-style order(Column major): [[0. 1. 2. 3. 4.] [5. 6. 7. 8. 9.]] 0.0,5.0,1.0,6.0,2.0,7.0,3.0,8.0,4.0,9.0, #22 matrix = np.arange(10, dtype=np.int8) print(matrix,len(matrix)) matrix = np.append(matrix,[11,22,33,44]) print(matrix,len(matrix)) o/p-[0 1 2 3 4 5 6 7 8 9] 10 [ 0 1 2 3 4 5 6 7 8 9 11 22 33 44] 14 #23 Tuple matrix = np.arange(10, dtype=np.int8) print(matrix,len(matrix)) matrix = np.append(matrix,[11,22,33,44]) print(matrix,len(matrix)) matrix = np.append(matrix, [55, 66, 77, 88]) print(matrix, len(matrix)) o/p-[0 1 2 3 4 5 6 7 8 9] 10 [ 0 1 2 3 4 5 6 7 8 9 11 22 33 44] 14 [ 0 1 2 3 4 5 6 7 8 9 11 22 33 44 55 66 77 88] 18 #24 matrix = np.arange(10, dtype=np.int8) print(matrix,len(matrix)) matrix = np.append(matrix,[11,22,33,44]) print(matrix,len(matrix)) matrix = np.append(matrix, [55, 66, 77, 88]) print(matrix, len(matrix)) matrix = np.delete(matrix,[5,6,8]) print(matrix) o/p-[0 1 2 3 4 5 6 7 8 9] 10 [ 0 1 2 3 4 5 6 7 8 9 11 22 33 44] 14 [ 0 1 2 3 4 5 6 7 8 9 11 22 33 44 55 66 77 88] 18 [ 0 1 2 3 4 7 9 11 22 33 44 55 66 77 88] #25 matrix = np.array([22,44,77,11,88,66]) print (matrix) matrix1=np.sort(matrix)#ascending order print(matrix1) matrix1 = -np.sort(-matrix)#descending order print(matrix1) o/p-[22 44 77 11 88 66] [11 22 44 66 77 88] [88 77 66 44 22 11] #26 matrix1=np.array([[11,22],[33,44]]) print(matrix1) matrix2=np.array([[55,66],[77,88]]) print((matrix2)) #joining these two matrices along axis =0 matrix3 = np.concatenate((matrix1,matrix2)) print(matrix3) o/p-[[11 22] [33 44]] [[55 66] [77 88]] [[11 22] [33 44] [55 66] [77 88]] #27 matrix1=np.array([[11,22],[33,44]]) print(matrix1) matrix2=np.array([[55,66],[77,88]]) print((matrix2)) #joining these two matrices along axis =0 matrix3 = np.concatenate((matrix1,matrix2)) print(matrix3) #joining these two matrices along axis =1 matrix4 = np.concatenate((matrix1,matrix2),axis=1) print(matrix4,matrix4.shape,matrix4.ndim) o/p-[[11 22] [33 44]] [[55 66] [77 88]] [[11 22] [33 44] [55 66] [77 88]] [[11 22 55 66] [33 44 77 88]] (2, 4) 2 #28 print("Stack the two arrays along axis 0") matrix3 =np.stack((matrix1,matrix2)) print(matrix3) matrix4= np.stack((matrix1,matrix2), axis=1) print(matrix4,matrix4.shape,matrix4.ndim)

o/p-Stack the two arrays along axis 0 [[[11 22] [33 44]]

[[55 66] [77 88]]] [[[11 22] [55 66]]

[[33 44] [77 88]]] (2, 2, 2) 3 #29 print("Stack the two arrays along axis 0") matrix3 = np.stack((matrix1, matrix2)) print(matrix3, matrix3.shape, matrix3.ndim) matrix4 = np.stack((matrix1, matrix2), axis = 1) print(matrix4, matrix4.shape, matrix4.ndim) o/p-Stack the two arrays along axis 0 [[[11 22] [33 44]]

[[55 66] [77 88]]] (2, 2, 2) 3 [[[11 22] [55 66]]

[[33 44] [77 88]]] (2, 2, 2) 3 #30 arr2=np.arange(20).reshape(4,5) print(arr2) print(arr2[2,:]) print(arr2[:,2]) print(arr2[1:3,2:4]) o/p-[[ 0 1 2 3 4] [ 5 6 7 8 9] [10 11 12 13 14] [15 16 17 18 19]] [10 11 12 13 14] [ 2 7 12 17] [[ 7 8] [12 13]] #31 arr2=np.arange(20).reshape(4,5) print(arr2) print(arr2[2,:]) print(arr2[:,2]) print(arr2[1:3,2:4]) print(arr2[1:,2:]) print(arr2[0::2]) print(arr2[0::2,1::2]) o/p-[[ 0 1 2 3 4] [ 5 6 7 8 9] [10 11 12 13 14] [15 16 17 18 19]] [10 11 12 13 14] [ 2 7 12 17] [[ 7 8] [12 13]] [[ 7 8 9] [12 13 14] [17 18 19]] [[ 0 1 2 3 4] [10 11 12 13 14]] [[ 1 3] [11 13]] #32 arr1 = np.arange(10) print(arr1,type(arr1)) arr_copy=arr1[2:5].copy() #creating a copy print(arr_copy) arr_copy[:]=100 print(arr1) print(arr_copy) o/p-[0 1 2 3 4 5 6 7 8 9] <class 'numpy.ndarray'> [2 3 4] [0 1 2 3 4 5 6 7 8 9] [100 100 100] #33 import numpy as np

arr = np.zeros((6, 6), dtype=int) arr[0, :] = 1 arr[-1, :] = 1 arr[:, 0] = 1 arr[:, -1] = 1 print(arr) o/p-[[1 1 1 1 1 1] [1 0 0 0 0 1] [1 0 0 0 0 1] [1 0 0 0 0 1] [1 0 0 0 0 1] [1 1 1 1 1 1]] or arr2=np.zeros((6,6)) arr2[[0,-1],:] = 1 arr2[:,[0,-1]] = 1 print(arr2) #34 arr1 = np.array([100, 20, -33, 44, 90, 35, -22]) print(arr1) print(arr1<0) print(arr1[arr1 < 0]) print((arr1 < 0)) print(arr1[(arr1 < 0)]) o/p-[100 20 -33 44 90 35 -22] [False False True False False False True] [-33 -22] [ True True False True True True False] [100 20 44 90 35] #35 arr1 = np.array([100, 20, -33, 44, 90, 35, -22]) print(arr1) print(arr1%2==True) print(arr1[arr1%2==0])#retrieving even numbers print(arr1[~(arr1%2==0)])#retrieving odd numbers o/p-[100 20 -33 44 90 35 -22] [False False True False False True False] [100 20 44 90 -22] [-33 35] #36 arr1 = np.array([100,20,-33,44,90,35,-22]) print(arr1) print((arr1>=20) & (arr1 <=50)) print(arr1[(arr1 >=20) & (arr1 <=50)]) print(np.version) print(arr1.nbytes) #retrieving numbers between 20 and 50 o/p-[100 20 -33 44 90 35 -22] [False True False True False True False] [20 44 35] 2.3.2 56 #37 arr1 = np.array(['bob','pat','tom','bob','jim','john']) print(arr1=='BOB') print(arr=='bob') print(arr1[arr1!='bob']) arr1[arr1=='bob'] = 'tim' print(arr1) #changing values o/p-[False False False False False False] [[False False False False False False] [False False False False False False] [False False False False False False] [False False False False False False] [False False False False False False] [False False False False False False]] ['pat' 'tom' 'jim' 'john'] ['tim' 'pat' 'tom' 'tim' 'jim' 'john'] #38 nums=np.arange(6) nums_sqrt = np.sqrt(nums) print(nums_sqrt) print(np.round(nums_sqrt,1),np.round(nums_sqrt,4)) print(np.ceil(nums_sqrt)) print(np.floor(nums_sqrt)) o/p-[0. 1. 1.41421356 1.73205081 2. 2.23606798] [0. 1. 1.4 1.7 2. 2.2] [0. 1. 1.4142 1.7321 2. 2.2361] [0. 1. 2. 2. 2. 3.] [0. 1. 1. 1. 2. 2.] #39 #dealing with random numbers nums = np.random.randn(10) print(nums) print(np.average(nums),np.std(nums)) o/p-[ 0.51421423 -0.9222735 0.10628887 -0.96270085 0.75370239 1.24193707 -1.56016315 0.42612416 1.13627468 0.18018588] 0.09135897871684981 0.8955100910703041 #40 nums = np.random.randn(500000) print(nums[:10]) print(np.average(nums),np.std(nums)) o/p-[ 1.58082713 1.04704363 -0.43345807 0.14128454 0.28660876 1.12530517 -0.35419236 1.56857163 -1.4136639 0.69523389] -0.002836505603716415 0.999861747073846 #41 np.random.seed(10) print(np.random.randn(10)) o/p-[ 1.3315865 0.71527897 -1.54540029 -0.00838385 0.62133597 -0.72008556 0.26551159 0.10854853 0.00429143 -0.17460021] #42 print(np.random.randint(0,5,20)) o/p-[3 2 1 0 4 1 3 3 1 4 1 4 1 1 4 3 2 0 3 4] #43 pt1=np.random.randn(10) pt2=np.random.randn(10) e_dist =0 for i in range(len(pt1)): e_dist += (pt1[i]-pt2[i])**2 print(e_dist **0.5) print (np.sqrt(np.sum((pt1-pt2)**2))) print(np.sqrt(np.sum(pow(np.subtract(pt1, pt2),2)))) o/p-1.9403940858656452 3.395792881516465 3.395792881516465 2.384539110887282 3.395792881516465 DAY_2 Data Gathering Data Cleaning/Santization Exploring the data analytics #45 import numpy as np
import pandas as pd

creating Series from a list

list_data = [101, 202, 404, 303, 505] print(list_data, type(list_data)) s = pd.Series(data=list_data, index=['red','green','brown','blue','magenta']) print(s) print(s[0], s['red'], s[4], s['magenta']) # indexing print(s[-1], s['magenta']) # indexing print(s[2:]) # slicing (2nd inde onwards) print(s[:-1]) # slicing (from first index till the last but one) o/p-[101, 202, 404, 303, 505] <class 'list'> red 101 green 202 brown 404 blue 303 magenta 505 dtype: int64 101 101 505 505 505 505 brown 404 blue 303 magenta 505 dtype: int64 red 101 green 202 brown 404 blue 303 dtype: int64 #46 print(s) print(s.sort_index()) print(s.sort_values()) print(s.argmin(), s.argmax(), len(s), s.count(), s.mean(), s.var(), s.std(), s.max(), s.min(), s.sum()) o/p-red 101 green 202 brown 404 blue 303 magenta 505 dtype: int64 blue 303 brown 404 green 202 magenta 505 red 101 dtype: int64 red 101 green 202 blue 303 brown 404 magenta 505 dtype: int64 0 4 5 5 303.0 25502.5 159.69502183850315 505 101 1515 #47 user_data = [['alice',19,'F','student'],['john',26,'M','student']] user_columns = ['name','age','gender','job'] user1 = pd.DataFrame(data=user_data, columns=user_columns) user1 o/p-user_data = [['alice',19,'F','student'],['john',26,'M','student']] user_columns = ['name','age','gender','job'] user1 = pd.DataFrame(data=user_data, columns=user_columns) user1 #48 user_data = dict(name=['eric','paul'], age=[22,58], gender=['M','F'], job=['student','manager']) print(user_data, type(user_data)) user2 = pd.DataFrame(data=user_data) user2 o/p- name age gender job 0 eric 22 M student 1 paul 58 F manager #49 user_data = {'name':['peter','paul'], 'age':[33,44], 'gender':['M','F'], 'job':['engineer','scientist']} print(user_data, type(user_data)) user3 = pd.DataFrame(data=user_data) user3 o/p- name age gender job 0 peter 33 M engineer 1 paul 44 F scientist #50 users = pd.concat([user1, user2, user3], ignore_index=True) print(users) o/p- name age gender job 0 alice 19 F student 1 john 26 M student 2 eric 22 M student 3 paul 58 F manager 4 peter 33 M engineer 5 paul 44 F scientist #51 arr_data=np.array(users) print(arr_data, type(arr_data)) arr_data=users.to_numpy() print(arr_data, type(arr_data)) o/p-[['alice' 19 'F' 'student'] ['john' 26 'M' 'student'] ['eric' 22 'M' 'student'] ['paul' 58 'F' 'manager'] ['peter' 33 'M' 'engineer'] ['paul' 44 'F' 'scientist']] <class 'numpy.ndarray'> [['alice' 19 'F' 'student'] ['john' 26 'M' 'student'] ['eric' 22 'M' 'student'] ['paul' 58 'F' 'manager'] ['peter' 33 'M' 'engineer'] ['paul' 44 'F' 'scientist']] <class 'numpy.ndarray'> #52 #join dataframes dict_data=dict(name=['alice','john','eric','paul','peter'], hieght=[165,180,175,180,181]) user5=pd.DataFrame(data=dict_data) user5 o/p- name hieght 0 alice 165 1 john 180 2 eric 175 3 paul 180 4 peter 181 #53 shape-no.of rows,no. of columns in form of tuples size-row * column len-no of rows users.shape[0]=rows users.shape[1]=columns