Scribd
Upload
11 views3 pages

3 Outliers Iqr

The document discusses outlier detection and removal using the Interquartile Range (IQR) method on a dataset of heights. It provides a step-by-step approach to calculate the IQR, identify outliers, and remove them from the dataset. Additionally, it includes an exercise for the reader to apply similar techniques on a different dataset containing heights and weights.

Uploaded by

kart238
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
11 views3 pages

3 Outliers Iqr

The document discusses outlier detection and removal using the Interquartile Range (IQR) method on a dataset of heights. It provides a step-by-step approach to calculate the IQR, identify outliers, and remove them from the dataset. Additionally, it includes an exercise for the reader to apply similar techniques on a different dataset containing heights and weights.

Uploaded by

kart238
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

3-outliers-iqr

August 13, 2024

Outlier Detection and Removal Using IQR


[3]: import pandas as pd
df = pd.read_csv("heights.csv")
df

[3]: name height


0 mohan 1.2
1 maria 2.3
2 sakib 4.9
3 tao 5.1
4 virat 5.2
5 khusbu 5.4
6 dmitry 5.5
7 selena 5.5
8 john 5.6
9 imran 5.6
10 jose 5.8
11 deepika 5.9
12 yoseph 6.0
13 binod 6.1
14 gulshan 6.2
15 johnson 6.5
16 donald 7.1
17 aamir 14.5
18 ken 23.2
19 Liu 40.2

[4]: df.describe()

[4]: height
count 20.000000
mean 8.390000
std 8.782812
min 1.200000
25% 5.350000
50% 5.700000
75% 6.275000

1
max 40.200000

Detect outliers using IQR


[5]: Q1 = df.height.quantile(0.25)
Q3 = df.height.quantile(0.75)
Q1, Q3

[5]: (5.3500000000000005, 6.275)

[6]: IQR = Q3 - Q1
IQR

[6]: 0.9249999999999998

[7]: lower_limit = Q1 - 1.5*IQR


upper_limit = Q3 + 1.5*IQR
lower_limit, upper_limit

[7]: (3.962500000000001, 7.6625)

Here are the outliers


[8]: df[(df.height<lower_limit)|(df.height>upper_limit)]

[8]: name height


0 mohan 1.2
1 maria 2.3
17 aamir 14.5
18 ken 23.2
19 Liu 40.2

Remove outliers
[9]: df_no_outlier = df[(df.height>lower_limit)&(df.height<upper_limit)]
df_no_outlier

[9]: name height


2 sakib 4.9
3 tao 5.1
4 virat 5.2
5 khusbu 5.4
6 dmitry 5.5
7 selena 5.5
8 john 5.6
9 imran 5.6
10 jose 5.8
11 deepika 5.9

2
12 yoseph 6.0
13 binod 6.1
14 gulshan 6.2
15 johnson 6.5
16 donald 7.1

Exercise
You are given height_weight.csv file which contains heights and weights of 1000 people. Dataset
is taken from here, https://www.kaggle.com/mustafaali96/weight-height
You need to do this,
(1) Load this csv in pandas dataframe and first plot histograms for height and weight parameters
(2) Using IQR detect weight outliers and print them
(3) Using IQR, detect height outliers and print them
Solution

You might also like