An Overview of Python for Data Analytics

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 12 | Dec 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 461
An Overview of Python for Data Analytics
Shailendra Chaudekar1
1PG Diploma, Big Data Analytics, CDAC Kolkata, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The data analysis in this research is done with
Python. The programming language is being explored. This
article quickly describes the fundamental processes of data
analysis, such as cleaning, transforming, and modelling data,
with an emphasis on exploratory data analysis. Data analysis
of an existing dataset and discovery of insights Using several
Python libraries and methods, some graphicalanalysisofdata
from the dataset will be demonstrated. Here, A dataset titled
"World Happiness Report 2022"isusedtoexamineandextract
data in both numerical and visual formats.
Key Words: Python, Data Analytics, Data Science, Data
Visualization, Python Libraries, Numpy, Pandas, Graphical
Exploratory Data Analytics
1. INTRODUCTION
Python is a high-level, general-purpose programming
language that has recently gained popularity. It enables
programmers to write code in fewer lines, which is not
achievable with other languages. Python programming is
notable for its support of many programming paradigms.
Python includes a broad, extendable standard library.
Python's primary characteristics include simple and easy to
learn, freeware and open source, and high-level
programming language. Platform independence,portability,
dynamically typed, procedure and object-oriented,
Extensible, Embedded, and Extensive Library.
We hope to provide a quick overview of Python in the areas
of data science, IoT, and machine learning in this article.
Python is well-known for having a plethora of libraries that
aid with data analysis and scientific computing.For example,
we may create a Python programmetoassistdata analystsin
analysing massive volumes of data for scientific computing.
This paper requires a basic understanding of statistics as
well as some experience with any C-style language. A
working understanding of Python is advantageous but not
required.
1.1 Introduction to Data Science
Data science is a multidisciplinary field that extracts
knowledge and insights from organised and unstructured
data using scientific techniques, procedures, tools, and
systems. Data analytics, data mining, and big data are all
connected to data science. It comprehends the data
phenomena. It incorporates techniques and theories from a
variety of domains, including mathematics, statistics,
computer science, and information science. Statistics is one
of the most essential disciplines for providing tools and
methods for finding structure in and providing greater
insight into data, as well as for analysing and quantifying
uncertainty. Python has a number of preconfiguredmodules
for working on data science tasks.
1.2 Python in Data Science Research and Education
Data Science is a booming field of study that combines
statistics, computer science, and a variety of applied
scientific fields. As is customary in such transdisciplinary
settings, fields of study, Data Science education,mentorship,
and research takes ideas and inspiration from a variety of
other domains, including the mathematical sciences,
computer science, and numerous business and application
domains.
This field presents an overview to Data Science, including
a wide range of essential issues and approaches for working
with huge data. Data collection, integration, management,
modelling, analysis, visualisation, prediction, and informed
decision-making are among the topics to be covered, as are
data security and data privacy. This integrates databases,
data warehousing, statistics, data mining, data visualisation,
high-performancecomputing,cloudcomputing,andbusiness
intelligence into the basic disciplines of data science.
Professional abilities such as communication, presentation,
and data storytelling will be cultivated. Through hands-on
projects and case studies in a range of business,engineering,
social sciences, and biological sciences areas, students will
get a practical grasp of data science
2. DATA ANALYSIS USING PYTHON
The analysis of varied data basically entails cleaning the
data, translating it into comprehensible form, and then
modelling data to extract some relevant information for
commercial or organisational usage.Itismostlyemployed in
commercial decisions. There are several libraries accessible
for conducting the analysis. NumPy, Pandas, Seaborn,
Matplotlib, Sklearn, and more are examples.
 NUMPY : NumPy is the foundational Python library
for scientific computing. It is a Python library that
includes a multidimensional array object, various
derived objects (such as masked arrays and
matrices), and a variety of routines for performing
fast array operations such as mathematical, logical,
shape manipulation, sorting, selecting, I/O,discrete
Fourier transforms, basic linear algebra, basic

statistical operations,randomsimulation,andmuch
more.
 PANDAS : Pandas is a data manipulation and
analysis software package created for the Python
computer language. It provides data structures and
functions for manipulating numerical tables and
time series in particular. It is free software
distributed under the BSD three-clause licence.The
name is taken from the word "panel data," which is
an econometrics term for data sets that comprise
observations for the same persons over several
periods. Its name is an allusion to the term "Python
data analysis." Wes McKinney began developing
what would become pandas while working as a
researcher at AQR Capital from 2007 to 2010.
 MatPlotlib : Matplotlib is a Python package that
allows you to create static, animated, and
interactive visualisations. Matplotlib makes simple
things simple and difficult things possible.
1. Produce plots suitable for publishing.
2. Create interactive figures that can be
zoomed, panned, and updated.
3. Change the visual style and layout.
4. Export to a variety of file formats.
5. Incorporate JupyterLaband Graphical User
Interfaces.
 Sklearn : The Sklearn Library ismostlyusedfordata
modelling andprovideseffective,user-friendlytools
for any type of predictive data analysis. This
library's principal use cases may be divided into six
groups, which are as follows:
1. Pre-processing
2. Regression
3. Classification
4. Clustering
5. Model Selection
6. Dimensionality Reduction
A. Data Requirements
The most crucial unit in every study is data. Data must be
given as inputs to the analysis based on the needs of the
analysis. The word "experimental unit" refers to the sort of
organisation employed to collect data (forexample,a person
or community of individuals). Specific demographic
variables (such as height, weight, age, and pay) can be
identified and obtained. It makes no difference whether the
data is numerical or categorical.
B. Data Collection:
Depending on the demands of the study, data is acquired
from a range of sources,includingrelational databases,cloud
databases, and other sources. Data sources can also include
field sensors such as traffic cameras, satellites, and
monitoring systems.
C. Data processing
Data must be processed or structured before it can be
analysed. These may include, for example, organising data
into rows and columns in a tabular format (known as
structured data) for further analysis, generally using a
spreadsheet or statistical software.
D. Data cleaning:
Data cleaning is the process of cleansing data after it has
been processed and arranged. It searches for and eliminates
data inconsistencies, duplication, and mistakes. Record
matching, identifying data inaccuracy, data sorting, outlier
data detection, textual data spell checking, and data quality
maintenance are all part of the data cleaning process. As a
result, it prevents unexpected consequences and aids us in
supplying high-quality data, which is critical for a successful
outcome.
E. Exploratory data analysis:
After the datasets have been cleaned and checked for errors,
they may be examined. A range of approaches, such as
exploratory data analysis understanding the messages
contained within the gathered data and descriptive
statistics—finding the average, median, and so on can be
used. Data visualisation is a technique in which data is
displayed graphically in order to get more insights into the
information contained within the data.
F. Data Modelling and algorithms:
Mathematical formulae or models can be applied to data to
detect links between variables, such as correlation or
causation.
G. Data product
A data product is a computer application that takes data
inputs and generates outputs that are then fed back into the
environment. It might be built on a model or an algorithm.
This review will explore data analysis in Python. The most
fundamental concepts, such as why Python is used for data
analysis, will be grasped. Furthermore, how anyone may
begin using Python will be demonstrated. The essential

libraries, platforms, and datasets for doing the analysis will
be introduced. The use of several Python functions for
numerical analysis is explained, as well as various ways for
generating graphs or charts.
Packages used: 1. Numpy 2.Pandas 3. Seaborn 4. Matplotlib
Platform used: Anaconda (Jupyter Notebook
Dataset used: World Happiness record 2022
Fig -1: A view of the dataset (World Happiness record2022)
A. Import libraries:
Libraries that will be utilised in the analysisprocess
should be imported first. Here are the import codes
for the libraries.
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline
import seaborn as sns
import sklearn
import csv
Fig -2: Importing libraries
B. Import Dataset
The dataset (World HappinessReport2022)isimportedinto
the Jupyter notebook in this case.
data = pd.read_csv("PathWorld HappinessReport2022.csv")
data
Fig -3: Importing dataset
C. Data Cleaning:
Data cleaning involves the removal of unnecessary data
or null values. So, initially, we must examine the dataset to
see whether it includes any null values or empty cells. #
isnull() returns true in entries with no value or a NA value.
And sum() is used in conjunction with isnull() to getthetotal
number of null values in each column.
data.isnull().sum()

Fig. 4. Checking null values in the dataset
We can extract specific rows or records from the dataset
based on our analytical needs. Here's an example of
extracting the first and end rows of a dataset.
#head() returns the data at the top of the dataset. Thehead's
default value is 5. (). The top ten rows of the datasetareused
in this case.
headdata = data.head(10)
headdata
Fig. 5. Top 10 rows of the dataset
#tail() is used to retrieve the dataset's final rows. The tail()
default value is 5.
taildata=data.tail(10)
Taildata
Fig. 6. Last 10 rows of the dataset
D. Exploratory Data Analysis In statistics:
Exploratory data analysis is a process of analysing
data sets in order to summarise their major
properties, which is frequently done withstatistical
graphics and other data visualisation approaches.A
statistical model may or may not be utilised, butthe
primary goal of EDA is to explore what the data can
tell us beyond the formal modelling or hypothesis
testing work. John Tukey championed exploratory
data analysis to encourage statisticians to
investigate the data and maybe create hypotheses
that could lead to future data gathering and
experiments.
Data types:
In Python, datatype refers to the type of data- int, object,and
float are the main datatypes. Using dtypes, print the data
types of all columns in the dataset.
data.dtypes
Fig. 7. Datatypes of the whole columns in the dataset

Describing the Dataset:
Describing data from a dataset entails obtaining a summary
of the supplied data frame, such as mean, count, min, max,
and so on. It is possible to accomplish this using the
describe() function-
data.describe()
Fig. 8. Summary of the whole dataset
taildata.describe()
Correlations:
Correlation demonstrates the relationship betweenanytwo
variables in a dataset. Correlation measuresthestrength ofa
linear relationship between two variables. Using corr() to
print the correlation of multiple characteristics
data.corr()
Fig. 10. Correlation of the whole dataset
data[['Country’, ‘Happiness score’, ‘GDP per capita', 'Social
support’, ‘Healthy life expectancy’, ‘Perceptions of
corruption']].corr()
corrdata
E. Graphical Exploratory Data Analysis
Graphical exploratory data analysis is
fundamentally the graphical version of non-
graphical exploratory data analysis. EDA evaluates
data sets to summarise their statistical properties
by focusing on the same four major elements, such
as measures of central tendency, measures of
spread, distribution shape, and the presence of
outliers. We also classified GEDA into three types:
univariate, bivariate, and multivariate. In the
coming paragraphs and GEDA features [5,] we will
go over these significant types in further depth. To
begin, a subset of the data frame is selected for
analysis or visualisation.
Fig. 9. Summary of some selected entries(10 last rows)
Fig. 11. Correlation of some attributes in the dataset

Fig. 12. A subset of the dataframe
F. Univariate Graphical Exploratory Data Analysis
1. Histogram:
A histogram is a data representation that resembles a bar
graph and groups various outcomes into columns along the
x-axis. The y-axis can be used to depict numerical counts or
percentages of occurrences in each column to demonstrate
data distributions. Matplotlib.pyplot.hist() can be used to
create a histogram in Python.
Chart. 1. Histogram
2. Stem Plot:
A stem plot establishes a marker at each x-point and draws
vertical lines from the baseline to the y-axis. The x-positions
are not required. The formats can be supplied askeyword or
positional parameters. Matplotlib.pyplot.stem canbeused to
create a stem plot in Python ()
Chart. 2. Stem plot
3. Box Plot:
A box plot is a visual depiction and comparison of data
groupings. The box plot illustrates the level, spread, and
symmetry of a data distribution by employing the median,
approximate quartiles, outliers, and the smallest and largest
data points (extreme values).
Chart. 3. Boxplot

Multivariate Graphical Exploratory Data Analysis
1. Scatter plot:
Chart. 4. Scatter Plot
• Heat Maps:
A heatmap is a graphical representationofdata thatemploys
colour coding to indicate different values. It is a two-
dimensional table of colour tones. This graphingapproachis
widely used in biology to display gene expression and other
multivariate data.
Chart. 5. Heatmap
2. Count Plot:
A Seaborn count plot is a graphical depiction of the number
of occurrences or frequency for eachcategoryofdata,shown
by bars. The countplot() function is used to generate bars
that represent the number of observations in each
categorical category. Thesub-data data frameisshownusing
the Count plot.
Chart. 6. Count plot
3. CONCLUSIONS
The many processes of data analysis, including data
collection, cleaning, and analysis, are addressed briefly in
this work. The primary focus of this course is exploratory
data analysis. Python programming language is utilised for
implementation. Jupyternotebook isusedforin-depthstudy.
Various Python libraries and packages are discussed.
Numerous findings are gatheredusingvariousanalytical and
visualization approaches. The dataset "World Happiness
Record 2022" is used to extract important information such
as the difference in happiness scores of different countries,
the dependence of one attribute in building up the score,
how a variable affects another variable, and so on. Various
graphs have been plotted using various attributes in the
dataset to draw conclusions in an easy way. The primary
focus of this course is exploratory data analysis. Python
programming language is utilised for implementation.
Jupyter notebook is used for in-depth study. Various Python
libraries and packages are discussed. Numerousfindings are
gathered using various analytical and visualization
approaches. The dataset "World Happiness Record 2022" is
used to extract important information such as thedifference
in happiness scores of different countries,thedependence of
one attribute in building up the score, how a variable affects
another variable, and so on. Various graphs have been
plotted using various attributes in the dataset to draw
conclusions in an easy way.
In a scatter plot, dots represent the values of two separate
numerical variables. The positions of each dot on the
horizontal and vertical axes represent the values for each
data point. Scatter plots are used to show how variables
relate to one another. The scatter plot of "Happiness score"
vs “GDP per Capita” is shown below-

REFERENCES
[1] Wolfram, S.: Mathematica: A System for Doing
Mathematics by Computer. Addi-son Wesley Longman
Publishing Co., Inc., Boston (1991)
[2] Mauriciusa Munhoz de Medeiros, Norberto Hoppen,
Antonio Carlosa Gastaud Maçada, Data science for
business: benefits,challengesandopportunities,Bottom
Line (ISSN: 0888045X) 33 (2020) 149–163,
http://dx.doi.org/10.1108/BL-12- 2019-
0132/FULL/XML.
[3] Viv Bewick, Liz Cheek, and Jonathan Ball. Statistics
review 7: Correlation and regression.Critical care,2003.
[4] Dr Ossama Embarak,Embarak,andKarkal.Data analysis
and visualization using python. Springer, 2018.
[5] Michel Jambu. Exploratory and multivariate data
analysis. Elsevier, 1991.
[6] Matthieu Komorowski, Dominic C Marshall, Justin D
Salciccioli, and Yves Crutain. Exploratory data analysis.
Secondary analysis of electronic health records, 2016.
[7] https://stackoverflow.com
[8] https://github.com
[9] KD Nuggets poll result https://www.kdnuggets.com/

An Overview of Python for Data Analytics

More Related Content

Similar to An Overview of Python for Data Analytics (20)

More from IRJET Journal (20)

Recently uploaded (20)

An Overview of Python for Data Analytics