DATA FRAMES

Dataframe objects of Pandas can store 2D heterogeneous data. It is a two-dimentional

data structure, just like any table ( with rows and columns )
# 1.CREATE DATA FRAME FROM LIST

import pandas as pd

stulist=[10,20,30,40,50]
stu_df1=pd.DataFrame(stulist)
print(stu_df1)
0
0 10
1 20
2 30
3 40
4 50
# 2. CREATE DATAFRAME FROM NESTED LIST

import pandas as pd
stulist=[['AAA','BBB','CCC','DDD','EEE','FFF'],[10,20,30,40,50,60]]
stu_df2=pd.DataFrame(stulist,index=['Name','Marks'],columns=['col1','col2','col3','col4','col5'
,'col6'])
print(stu_df2)
col1 col2 col3 col4 col5 col6
Name AAA BBB CCC DDD EEE FFF
Marks 10 20 30 40 50 60

# 3. CREATE DATAFRAME USING SERIES

#Create a Series using a Dictionary

stu_marks=pd.Series({'Anil':30,'Vishal':89,'Komal':67,'Varsha':35,'Pulkit':77})
stu_class=pd.Series({'Anil':11,'Vishal':12,'Komal':10,'Varsha':12,'Pulkit':11})
stu_df3=pd.DataFrame({'Marks':stu_marks,'Class':stu_class})
print(stu_df3)

Marks Class
Anil 30 11
Vishal 89 12
Komal 67 10
Varsha 35 12
Pulkit 77 11

#Assignment

Iname Price Company

I1 TV 4000.0 NaN
I2 AC 5000.0 SAMSUNG
I3 OVEN 9000.0 USHA
 I1 NaN NaN LG

#4. Create a dataframe with following structure:

# Iname Price Company

# I1 TV 4000 LG
# I2 AC 5000 SAMSUNG
# I3 OVEN 9000 USHA

#Solution to assignment

s1=pd.Series({'I1':'TV','I2':'AC','I3':'OVEN'})
s2=pd.Series({'I1':4000,'I2':5000,'I3':9000})
s3=pd.Series({'i1':'LG','I2':'SAMSUNG','I3':'USHA'})

idf=pd.DataFrame({'Iname':s1,'Price':s2,'Company':s3})
print(idf)

# 5.CREATE DATAFRAME from DICTIONARY with a list associated as the value for a key
import pandas as pd
import numpy as np
dict_stu=
{'Rollno':[1,2,3,4,5],'Name':['Anil','Ankit','Vishal','Gaurav','Harsh'],'Class':[11,11,12,10,12],
'Sec':['A','B','B','A','C']}
print("="*30)
studf=pd.DataFrame(dict_stu)
print(studf)
Roll no Name Class Sec
0 1 Anil 11 A
1 2 Ankit 11 B
2 3 Vishal 12 B
3 4 Gaurav 10 A
4 5 Harsh 12 C

# 6. Create a dataframe from an existing dataframe

import pandas as pd
studf1=pd.DataFrame(studf)
print(studf1)
# 7. Creating an indexed dataframe using lists
stu={'ID':['E1','E2','E3','E4','E5'],'RollNo':[1,2,3,4,5],'Name':
['Anil','Ankit','Vishal','Gaurav','Harsh'],'Class':[11,11,12,10,12],'Sec':['A','B','B','A','C']}
stdf=pd.DataFrame(stu, index=['S1','S2','S3','S4','S5'])
print(stdf)
ID RollNo Name Class Sec
S1 E1 1 Anil 11 A
S2 E2 2 Ankit 11 B
S3 E3 3 Vishal 12 B
S4 E4 4 Gaurav 10 A
S5 E5 5 Harsh 12 C

#8. Accessing elements of a dataframe

print(stdf['RollNo'])
print(stdf.RollNo)
S1 1
S2 2
S3 3
S4 4
S5 5
 #9. To access more than 1 column pass them as a list
print(stdf[['RollNo','Class']])
Name: RollNo, dtype: int64
S1 1
S2 2
S3 3
S4 4
S5 5
#10. Change the index by taking value from one of the column of dataframe

#stdf1=pd.DataFrame(stdf)
#stdf1.set_index('ID',inplace=True)
#print(stdf1)
ID RollNo Name Class Sec Marks
S1 E1 1 Anil 11 A 90
S2 E2 2 Ankit 11 B 90
S3 E3 3 Vishal 12 B 90
S4 E4 4 Gaurav 10 A 90
S5 E5 5 Harsh 12 C 90
#11. To add a new column in data frame with scalar value
stdf['Marks']=90
print(stdf)
 ID RollNo Name Class Sec Marks
S1 E1 1 Anil 11 A 90
S2 E2 2 Ankit 11 B 90
S3 E3 3 Vishal 12 B 90
S4 E4 4 Gaurav 10 A 90
S5 E5 5 Harsh 12 C 90

#12. Add a new column with varying values for each index value
stdf['Marks']=pd.Series([10,20,30,40,50],index=['S1','S2','S3','S4','S5'])
print(stdf)
ID RollNo Name Class Sec Marks
S1 E1 1 Anil 11 A 10
S2 E2 2 Ankit 11 B 20
S3 E3 3 Vishal 12 B 30
S4 E4 4 Gaurav 10 A 40
S5 E5 5 Harsh 12 C 50
stdf['New Marks']=stdf['Marks']+5
print(stdf)
ID RollNo Name Class Sec Marks New Marks
S1 E1 1 Anil 11 A 10 15
S2 E2 2 Ankit 11 B 20 25
S3 E3 3 Vishal 12 B 30 35
S4 E4 4 Gaurav 10 A 40 45
 S5 E5 5 Harsh 12 C 50 55

#13. Applying an expression with character values

stdf['NN']=stdf['Sec']+'Q'
stdf['New_Sec']=stdf['Sec']+"$"
print(stdf)

#Applying * operator to replicate the character values of a column

stdf['New Sec2']=stdf['Sec']*3
print(stdf)

stdf['New Sec2']=stdf['Sec']*3+"@"+"Name"
print(stdf)

stdf.drop(columns=['NN'],inplace=True)
print(stdf)

#Extracting data from dataframe (loc & iloc method)

#Syntax of loc method <dataframe>.loc(<start row heading>:<stop row heading>,<start
column heading<:<stop column heading>)
print(stdf.loc['S1':'S4','ID':'Class'])
#If row label and column heading is skipped it takes values
print(stdf.loc['S1':,:'Sec'])

#Syntax of iloc method:

# <dataframe>.iloc(<start row index>:<stop row index>,<start column index>:<stop column
index>)

print(stdf.iloc[0:2,0:2])
print(stdf.iloc[::-1])

print(stdf.iloc[::-1,::-1])

#AGGREGATION USING STATISTICAL FUNCTIONS ON DATAFRAMES

#Create dataframe of players with their score
Dfgame=pd.DataFrame({'Pno':[1,2,3,4,5,6,7,8],'Pname':
['Virat','Sehwag','Dhoni','Rahane','Hardik','Jadeja','Shami','Buruah'],'Score':
[90,98,86,76,70,65,93,80]})
print(Dfgame)

#Max() function

print("The highest score is:",Dfgame['Score'].max())

print("Max function applied on Pname function:",Dfgame['Pname'].max())

#Min function
print("the lowest score is:",Dfgame['Score'].min())
print("Min function applied on Pname function:",Dfgame['Pname'].min())

#Function Sum
print("The total score of team is :",Dfgame['Score'].sum())

#Count the total number of columns for each row

print("the total number of columns for each row is:")

print(Dfgame.count(axis=1))

==============================
Roll no Name Class Sec
0 1 Anil 11 A
1 2 Ankit 11 B
2 3 Vishal 12 B
3 4 Gaurav 10 A
4 5 Harsh 12 C
Roll no Name Class Sec
0 1 Anil 11 A
1 2 Ankit 11 B
2 3 Vishal 12 B
3 4 Gaurav 10 A
4 5 Harsh 12 C
 ID RollNo Name Class Sec
S1 E1 1 Anil 11 A
S2 E2 2 Ankit 11 B
S3 E3 3 Vishal 12 B
S4 E4 4 Gaurav 10 A
S5 E5 5 Harsh 12 C
S1 1
S2 2
S3 3
S4 4
S5 5
Name: RollNo, dtype: int64
S1 1
S2 2
S3 3
S4 4
S5 5
Name: RollNo, dtype: int64
RollNo Class
S1 1 11
S2 2 11
S3 3 12
S4 4 10
S5 5 12
 ID RollNo Name Class Sec Marks
S1 E1 1 Anil 11 A 90
S2 E2 2 Ankit 11 B 90
S3 E3 3 Vishal 12 B 90
S4 E4 4 Gaurav 10 A 90
S5 E5 5 Harsh 12 C 90
ID RollNo Name Class Sec Marks
S1 E1 1 Anil 11 A 10
S2 E2 2 Ankit 11 B 20
S3 E3 3 Vishal 12 B 30
S4 E4 4 Gaurav 10 A 40
S5 E5 5 Harsh 12 C 50
ID RollNo Name Class Sec Marks New Marks
S1 E1 1 Anil 11 A 10 15
S2 E2 2 Ankit 11 B 20 25
S3 E3 3 Vishal 12 B 30 35
S4 E4 4 Gaurav 10 A 40 45
S5 E5 5 Harsh 12 C 50 55
ID RollNo Name Class Sec Marks New Marks NN New_Sec
S1 E1 1 Anil 11 A 10 15 AQ A$
S2 E2 2 Ankit 11 B 20 25 BQ B$
S3 E3 3 Vishal 12 B 30 35 BQ B$
S4 E4 4 Gaurav 10 A 40 45 AQ A$
S5 E5 5 Harsh 12 C 50 55 CQ C$
 ID RollNo Name Class Sec Marks New Marks NN New_Sec New Sec2
S1 E1 1 Anil 11 A 10 15 AQ A$ AAA
S2 E2 2 Ankit 11 B 20 25 BQ B$ BBB
S3 E3 3 Vishal 12 B 30 35 BQ B$ BBB
S4 E4 4 Gaurav 10 A 40 45 AQ A$ AAA
S5 E5 5 Harsh 12 C 50 55 CQ C$ CCC
ID RollNo Name Class Sec Marks New Marks NN New_Sec New Sec2
S1 E1 1 Anil 11 A 10 15 AQ A$ AAA@Name
S2 E2 2 Ankit 11 B 20 25 BQ B$ BBB@Name
S3 E3 3 Vishal 12 B 30 35 BQ B$ BBB@Name
S4 E4 4 Gaurav 10 A 40 45 AQ A$ AAA@Name
S5 E5 5 Harsh 12 C 50 55 CQ C$ CCC@Name

ID RollNo Name Class Sec Marks New Marks New_Sec New Sec2
S1 E1 1 Anil 11 A 10 15 A$ AAA@Name
S2 E2 2 Ankit 11 B 20 25 B$ BBB@Name
S3 E3 3 Vishal 12 B 30 35 B$ BBB@Name
S4 E4 4 Gaurav 10 A 40 45 A$ AAA@Name
S5 E5 5 Harsh 12 C 50 55 C$ CCC@Name
ID RollNo Name Class
S1 E1 1 Anil 11
S2 E2 2 Ankit 11
S3 E3 3 Vishal 12
S4 E4 4 Gaurav 10
 ID RollNo Name Class Sec
S1 E1 1 Anil 11 A
S2 E2 2 Ankit 11 B
S3 E3 3 Vishal 12 B
S4 E4 4 Gaurav 10 A
S5 E5 5 Harsh 12 C
ID RollNo
S1 E1 1
S2 E2 2
ID RollNo Name Class Sec Marks New Marks New_Sec New Sec2
S5 E5 5 Harsh 12 C 50 55 C$ CCC@Name
S4 E4 4 Gaurav 10 A 40 45 A$ AAA@Name
S3 E3 3 Vishal 12 B 30 35 B$ BBB@Name
S2 E2 2 Ankit 11 B 20 25 B$ BBB@Name
S1 E1 1 Anil 11 A 10 15 A$ AAA@Name
New Sec2 New_Sec New Marks Marks Sec Class Name RollNo ID
S5 CCC@Name C$ 55 50 C 12 Harsh 5 E5
S4 AAA@Name A$ 45 40 A 10 Gaurav 4 E4
S3 BBB@Name B$ 35 30 B 12 Vishal 3 E3
S2 BBB@Name B$ 25 20 B 11 Ankit 2 E2
S1 AAA@Name A$ 15 10 A 11 Anil 1 E1
Pno Pname Score
0 1 Virat 90
1 2 Sehwag 98
2 3 Dhoni 86
3 4 Rahane 76
4 5 Hardik 70
5 6 Jadeja 65
6 7 Shami 93
7 8 Buruah 80
The highest score is: 98
Max function applied on Pname function: Virat
the lowest score is: 65
Min function applied on Pname function: Buruah
The total score of team is : 658
the total number of columns for each row is:
0 3
1 3
2 3
3 3
4 3
5 3
6 3
7 3
dtype: int64
>>>