Change data type using pandas

Let's see the dataset. This dataset is will be used in upcoming example of code .

Input
import pandas as pd
df=pd.read_csv("practice.csv")
print(df)
Output

	Id	Name	Group_name	Total_marks	Grade	Ranking
0	01	A	Science	700	A+	01
1	02	B	Commerce	618	B+	02
2	03	A	Science	700	A+	01
3	04	D	Arts	687	A+	01
4	05	E	Commerce	611	B+	02
5	06	F	Arts	599	C+	03
6	07	P	Science	575	C+	03
7	08	F	Arts	600	C	03
8	09	I	Commerce	550	C+	03
9	10	J	Science	650	A+	01
10	11	K	Arts	680	A+	01
11	12	L	Science	570	C+	03
12	13	M	Arts	599	C+	03
13	14	N	Commerce	597	C+	03
14	15	O	Science	697	A+	01
15	16	B	Arts	570	C+	03
16	17	D	Science	588	C+	03
17	18	E	Science	687	A+	01
18	19	C	Commerce	688	A+	01
19	20	P	Arts	588	C+	03
20	21	C	Science	619	B+	02
21	22	M	Commerce	600	B+	02
22	23	P	Arts	700	A+	01

astype function()

Astype function is used to data type conversion like float to int or string, string to float or int and int to float or string.

Input
import pandas as pd
df=pd.read_csv("practice.csv")

Print("Original Dtypes")
print(df.dtypes)

df["Id"]=df["Id"].astype(float)
df["Ranking"]=df["Ranking"].astype(int)
df["Total marks"]=df["Total marks"].astype(str)
df['Date'] = df['Date'].astype('datetime64[ns]')

print("Dtype after conversion")
print("New Id column",df["Id"].dtypes)
print("New Ranking column",df["Ranking"].dtypes)
print("New Total marks column",df["Id"].dtypes)
Output
Original Dtypes
Id    int64
Name    object
Group_name    object
Total marks    int64
Grade    object
Ranking    int64
Date    object
dtype: object

Dtype after conversion

New Id column float64
New Ranking column int32
New Total marks column float64

Float to integer tricks

To convert float to integer, round those float values before converting into integer.

Input
import pandas as pd
df=pd.read_csv("prac.csv")
print(df)
df["Ranking"]=df["Ranking"].round(0).astype(int)
print("New Ranking column",df["Ranking"].dtypes)
Output

	Id	Money
0	1.12	$4,334.00
1	2.31	$4,343.00
2	3.00	$4,352.00
3	4.11	$4,361.00
4	5.01	$4,370.00
5	6.21	$4,379.00

to_numeric function()

If you have string in numeric column then you can't convert that column into integer or float using astype. So you have to use to_numeric function and use an extra parameter and that is errors="coerce". When this parameter gets a string it will make that cell nan. After doing this you will get a numeric column and to change it into float or int you can use astype function.

Input
import pandas as pd
df=pd.read_csv("prac.csv")
df["Ranking"]=pd.to_numeric(df["Ranking"],errors="coerce").astype(int)
print("New Ranking column",df["Ranking"].dtypes)

Suppose you have some symbol(dollar, comma, etc) with numeric values. You can't convert those values into numeric from with those symbols. For the conversion, you have to remove those symbols. After removing the symbol you can change those values into float or integer.

Input
import pandas as pd
df=pd.read_csv("prac.csv")
print(df)
df["money"]=df["money"].str.replace("$","").str.replace(",","")
print(df)
df["money"].astype(float)
print("After changing dtypes",df["money"].dtypes)
Output

	Id	Money
0	01	$4,334.00
1	02	$4,343.00
2	03	$4,352.00
3	04	$4,361.00
4	05	$4,370.00
5	06	$4,379.00

Name: money, dtype: object

	Id	Money
0	01	4334.0
1	02	4343.0
2	03	4352.0
3	04	4361.0
4	05	4370.0
5	06	4379.0

Change date-time column data type

CodersAim is created for learning and training a self learner to become a professional from beginner. While using CodersAim, you agree to have read and accepted our terms of use, privacy policy, Contact Us