JIYIK CN >

This article will show you how to get the unique values ​​in a Pandas DataFrame column.

For example, suppose we have a DataFrame consisting of individuals and their occupations, and we want to know the total number of occupations. In this case, we cannot simply use the total number of rows to determine the total number of unique occupations, as many people may have the same job. For this case, we can use the unique()sum drop_duplicates()function provided by the Pandas library.

It is also important to know how to sort a DataFrame as it can help visualize and understand the data. The sorted()and sort_values()functions can help with this.

We will sort and delete the following DataFrame in this tutorial.

import pandas as pd
import numpy as np

df = pd.DataFrame({"A": [7, 1, 5, 4, 2, 1, 4, 4, 8], "B": [1, 2, 8, 5, 3, 4, 2, 6, 8]})

print(df)

Output:

   A  B
0  7  1
1  1  2
2  5  8
3  4  5
4  2  3
5  1  4
6  4  2
7  4  6
8  8  8

uniqueGet unique values ​​in a Pandas DataFrame column using

Pandas Series’s unique()get method is used when we process a single column of a DataFrame and returns all the unique elements of a column. unique()The final output of using the get function is an array.

example:

import pandas as pd
import numpy as np

df = pd.DataFrame({"A": [7, 1, 5, 4, 2, 1, 4, 4, 8], "B": [1, 2, 8, 5, 3, 4, 2, 6, 8]})

print(df["A"].unique())
print(type(df["A"].unique()))

Output:

[7 1 5 4 2 8]
numpy.ndarray

drop_duplicatesGet unique values ​​in a Pandas DataFrame column using

drop_duplicates()Can be applied to a DataFrame or a subset thereof and preserves the type of the DataFrame object. It is also considered a faster option when processing huge datasets to remove duplicate values.

example:

import pandas as pd
import numpy as np

df = pd.DataFrame({"A": [7, 1, 5, 4, 2, 1, 4, 4, 8], "B": [1, 2, 8, 5, 3, 4, 2, 6, 8]})

print(df.drop_duplicates(subset="A"))
print(type(df.drop_duplicates(subset="A")))

Output:

   A  B
0  7  1
1  1  2
2  5  8
3  4  5
4  2  3
8  8  8
pandas.core.frame.DataFrame

Sorting columns in a Pandas DataFrame

We can sorted()sort a column using the method, but it converts the final result into a list type object. We can also sort the column values ​​in descending order by reversedsetting the parameter to .True

The following example sorts a column in ascending order and removes duplicate values.

import pandas as pd
import numpy as np

df = pd.DataFrame({"A": [7, 1, 5, 4, 2, 1, 4, 4, 8], "B": [1, 2, 8, 5, 3, 4, 2, 6, 8]})

df_new = df.drop_duplicates(subset="A")

print(sorted(df_new["A"]))
print(type(sorted(df_new["A"])))

Output:

[1, 2, 4, 5, 7, 8]
list

sort_values()is another flexible option to sort a DataFrame. Here, we can byspecify the column to be sorted using the argument and ascendingwhether the order should be ascending or descending using the argument. It retains the object type as Pandas DataFrame.

The following example sorts the column in descending order and removes duplicate values.

import pandas as pd
import numpy as np

df = pd.DataFrame({"A": [7, 1, 5, 4, 2, 1, 4, 4, 8], "B": [1, 2, 8, 5, 3, 4, 2, 6, 8]})

df_new = df.drop_duplicates(subset="A")

print(df_new.sort_values(by="A", ascending=False))
type(df_new.sort_values(by="A"))

Output:

   A  B
8  8  8
0  7  1
2  5  8
3  4  5
4  2  3
1  1  2
pandas.core.frame.DataFrame