Get unique values from a column in Pandas DataFrame

You can get one-of-a-kind qualities in segments (different sections) from pandas DataFrame utilizing unique() or Series.unique() capabilities, unique() from Series is utilized to get one-of-a-kind qualities from a solitary segment, and the other one is utilized to get from numerous sections.

Pandas is a Python package that provides rapid, adaptive, and expressive data structures designed to make dealing with "social" or "named" data simple and natural. It plans to be the major significant level structure block for doing viable, genuine information examination in Python. It also has the broader goal of being the most powerful and versatile open-source information examination/control device available in any language. It is already well on its way to achieving this goal.

Fundamental Features

Here are just a few examples of how well pandas get along:

• Simple handling of missing information (addressed as NaN, NA, or NaT) in drifting point and non-drifting point data. Sections can be embedded and removed from DataFrame and higher layered objects due to size impermanence.
• Programmed and unambiguous data arrangement: items can be explicitly modified to a set of marks, or the client can completely disregard the labels and allow Series, DataFrame, and so on to alter the data for you in calculations.
• Strong, adaptable information gathering by usefulness to do split-apply-join method on informative indexes for both collecting and updating information.
• Make it straightforward to convert damaged, unexpectedly listed data from various Python and NumPy data structures into DataFrame objects.
• Astute name-based pruning, lavish sorting, and subsetting of massive informative indexes.Natural blending and joining informational collections.
• Adaptable reshaping and turning of informational collections.
• Tomahawk naming progression (conceivable to have different marks per tick)
• Powerful IO instruments for stacking data from CSV and delimited level documents, Excel records, data sets, and saving/stacking data from the ultrafast HDF5 design.
• Time series-specific applications include: date range age and recurrence transformation, moving window measurements, date shifting and slacking.

Example

import pandas as pd
import numpy as np
technology = {
    'Subjects':["C","C++","Python","pandas","Python","C","pandas"],
    'Fee' :[20000,25000,22000,30000,22000,20000,30000],
    'Duration':['30days','40days','35days','50days','40days','30days','50days'],
    'Offer':[1000,2300,1200,2000,2300,1000,2000]
              }
df = pd.DataFrame(technology)
print(df)

Output

pandas Get Unique Values in Column

Exceptional is likewise alluded to as particular; you can get special qualities in the segment utilizing the pandas Series. Unique () capability, since this capability needs to approach the Series object, use df['column_name'] to get the extraordinary qualities as a Series.

Syntax:

Series.unique(values)

Example

import pandas as pd
import numpy as np
technology = {
    'Subjects':["C","C++","Python","pandas","Python","C","pandas"],
    'Fee' :[20000,25000,22000,30000,22000,20000,30000],
    'Duration':['30days','40days','35days','50days','40days','30days','50days'],
    'Offer':[1000,2300,1200,2000,2300,1000,2000]
              }
df = pd.DataFrame(technology)
print(df['Subjects'].unique())

Output

Yields Series object as result. This kills all copies and returns just exceptional qualities from the Subjects section.

Finding Unique Values in Multiple Columns

In the case, you need to get one-of-a-kind qualities on numerous sections of DataFrame, use pandas.unique() capability, utilizing this, you can get special upsides of a solitary segment.

Syntax:

pandas.unique(values)

Example

<!-- wp:paragraph -->
import pandas as pd
import numpy as np
technology = {
    'Subjects':["C","C++","Python","pandas","Python","C","pandas"],
    'Fee' :[20000,25000,22000,30000,22000,20000,30000],
    'Duration':['30days','40days','35days','50days','40days','30days','50days'],
    'Offer':[1000,2300,1200,2000,2300,1000,2000]
              }
df = pd.DataFrame(technology)
df2 = pd.unique(df[['Subjects', 'Fee']].values.ravel())
print(df2)

Output

To get all special qualities for one segment and afterwards, the subsequent section use contention 'K' to the ravel() capability. The contention 'K' advises the technique to straighten the cluster in the request for the components. This can be altogether quicker than utilizing the technique's default 'C' request.

Example

import pandas as pd
import numpy as np
technology = {
    'Subjects':["C","C++","Python","pandas","Python","C","pandas"],
    'Fee' :[20000,25000,22000,30000,22000,20000,30000],
    'Duration':['30days','40days','35days','50days','40days','30days','50days'],
    'Offer':[1000,2300,1200,2000,2300,1000,2000]
              }
df = pd.DataFrame(technology)
df2 = pd.unique(df[['Subjects', 'Fee']].values.ravel('k'))
print(df2)

Output

Using Numpy.unique()

Assuming that you are utilizing Numpy, utilize remarkable() technique to dispense with copy values.

Example

import numpy as np
# Find the unique values in multiple columns using numpy.unique()
df2 = np.unique(df[['Subjects', 'Duration']].values)
print(df2)
# Use numpy.unique() to unique values in multiple columns 
column_values = df[['Subjects', 'Duration']].values
df2 = np.unique(column_values)
print(df2)

Output

Using set() to Eliminate Duplicates

The set() capability eliminates every copy esteem and gets just novel qualities. We can utilize this set() capability to get extraordinary qualities from DataFrame single or different sections.

Example

# Using Set() in pandas DataFrame
df2 = set(df.Subjects.append(df.Fee).values)
print(df2)
# Using set() method
df2 = set(df.Subjects) | set(df.Fee)
print(df2)

Output

Use pandas.concat() and Unique() Methods

Utilizing remarkable() and pandas.concat() mix to get interesting upsides of numerous segments.

# Using pandas.concat to extend one column to multiple columns
df2 = pd.concat([df['Subjects'],df['Duration'],df['Fee']]).unique()
print(f"Unique Values from three Columns: {df2}")

Output

Use Series.drop_duplicates()

At last, you can get the exceptional upsides of a segment utilizing the drop_duplicates() capability of the Series object. Subsequent to dropping copies, it returns a Series object with special qualities.

import pandas as pd
import numpy as np
technology = {
    'Subjects':["C","C++","Python","pandas","Python","C","pandas"],
    'Fee' :[20000,25000,22000,30000,22000,20000,30000],
    'Duration':['30days','40days','35days','50days','40days','30days','50days'],
    'Offer':[1000,2300,1200,2000,2300,1000,2000]
              }
df = pd.DataFrame(technology)
# Use Series.drop_duplicates() to get unique values
print(df.Subjects.drop_duplicates())

Output

Overall Example of pandas Get Unique Values in Columns

import pandas as pd
import numpy as np
technology = {
    'Subjects':["C","C++","Python","pandas","Python","C","pandas"],
    'Fee' :[20000,25000,22000,30000,22000,20000,30000],
    'Duration':['30days','40days','35days','50days','40days','30days','50days'],
    'Offer':[1000,2300,1200,2000,2300,1000,2000]
              }
df = pd.DataFrame(technology)
print(df)


# Find unique values of a column
print(df['Subjects'].unique())
print(df.Subjects.unique())


# Convert to List
print(df.Subjects.unique().tolist())


# unique values with drop_duplicates
df.Subjects.drop_duplicates()
print(df)


# Using pandas.unique() to unique values in multiple columns
df2 = pd.unique(df[['Subjects', 'Fee']].values.ravel('K'))
print(df2)


# Using pandas.unique() to unique values
df2 = pd.unique(df[['Subjects']].values.ravel())
print(df2)


# Find the unique values in multiple columns using numpy.unique()
df2 = np.unique(df[['Subjects', 'Duration']].values)
print(df2)


# Use numpy.unique() to unique values in multiple columns 
column_values = df[['Subjects', 'Duration']].values
df2 = np.unique(column_values)
print(df2)


# Using Set() in pandas DataFrame
df2 = set(df.Subjects.append(df.Fee).values)
print(df2)


# Using set() method
df2 = set(df.Subjects) | set(df.Fee)
print(df2)


# To get unique values in one series/column
df2 = df['Subjects'].unique()
print(df2)


# Using pandas.concat to extend one column to multiple columns
df2 = pd.concat([df['Subjects'],df['Duration'],df['Fee']]).unique()
print(df2)


# Use Series.drop_duplicates() to get unique values
print(df.Subjects.drop_duplicates())

Output: