EDA-Unit 1.pdf

EXPLORATORY DATA
ANALYSIS
DR. V. NIRMALA
DEPARTMENT OF AI & DS
EASWARI ENGINEERING COLLEGE

Introduction..
• Data encompasses a collection of discrete objects, numbers, words, events, facts, measurements,
observations, or even descriptions of things.
• Such data is collected and stored by every event or process occurring in several disciplines, including
biology, economics, engineering, marketing, and others.
• Processing such data elicits useful information and processing such information generates useful
knowledge. But an important question is:
• how can we generate meaningful and useful information from such data? An answer to this question is
EDA.
• EDA is a process of examining the available dataset to discover patterns, spot anomalies, test
hypotheses, and check assumptions using statistical measures.

Stages of EDA
Let's understand in brief what these stages are:
 Data requirements:
There can be various sources of data for an organization. It is important to comprehend what type of
data is required for the organization to be collected, curated, and stored. For example, an application tracking
the sleeping pattern of patients suffering from dementia requires several types of sensors' data storage, such
as sleep data, heart rate from the patient, electro-dermal activities, and user activities pattern. All of these
data points are required to correctly diagnose the mental state of the person. Hence, these are mandatory
requirements for the application. In addition to this, it is required to categorize the data, numerical or
categorical, and the format of storage and dissemination.
 Data collection:
Data collected from several sources must be stored in the correct format and transferred to the right
information technology personnel within a company. As mentioned previously, data can be collected from
several objects on several events using different types of sensors and storage tools.

Stages of EDA..
 Data processing:
Preprocessing involves the process of pre-curating the dataset before actual analysis. Common tasks
involve correctly exporting the dataset, placing them under the right tables, structuring them, and exporting
them in the correct format.
 Data cleaning:
Preprocessed data is still not ready for detailed analysis. It must be correctly transformed for an
incompleteness check, duplicates check, error check, and missing value check. These tasks are performed in
the data cleaning stage, which involves responsibilities such as matching the correct record, finding
inaccuracies in the dataset, understanding the overall data quality, removing duplicate items, and filling in the
missing values. However, how could we identify these anomalies on any dataset? Finding such data issues
requires us to perform some analytical techniques. We will be learning several such analytical techniques in
Data Transformation. To understand briefly, data cleaning is dependent on the types of data under study.
Hence, it is most essential for data scientists or EDA experts to comprehend different types of datasets. An
example of data cleaning would be using outlier detection methods for quantitative data cleaning.

Stages of EDA..
 EDA:
Exploratory data analysis, as mentioned before, is the stage where we actually start to understand the message contained in the
data. It should be noted that several types of data transformation techniques might be required during the process of exploration.
We will cover descriptive statistics in-depth in Section 2, Chapter 5, Descriptive Statistics, to understand the mathematical
foundation behind descriptive statistics. This entire book is dedicated to tasks involved in exploratory data analysis.
 Modeling and algorithm:
From a data science perspective, generalized models or mathematical formulas can represent or exhibit relationships among
different variables, such as correlation or causation. These models or equations involve one or more variables that depend on other
variables to cause an event. For example, when buying, say, pens, the total price of pens(Total) = price for one pen(UnitPrice) * the
number of pens bought (Quantity). Hence, our model would be Total = UnitPrice * Quantity. Here, the total price is dependent on
the unit price. Hence, the total price is referred to as the dependent variable and the unit price is referred to as an independent
variable.
In general, a model always describes the relationship between independent and dependent variables. Inferential statistics
deals with quantifying relationships between particular variables. The Judd model for describing the relationship between data,
model, and error still holds true: Data = Model + Error.

Stages of EDA..
 Data Product:
Any computer software that uses data as inputs, produces outputs, and provides feedback based
on the output to control the environment is referred to as a data product. A data product is generally
based on a model developed during data analysis, for example, a recommendation model that inputs
user purchase history and recommends a related item that the user is highly likely to buy.
 Communication:
This stage deals with disseminating the results to end stakeholders to use the result for business
intelligence. One of the most notable steps in this stage is data visualization. Visualization deals with
information relay techniques such as tables, charts, summary diagrams, and bar charts to show the
analyzed result.

The significance of EDA
• Different fields of science, economics, engineering, and marketing accumulate and store data primarily in
electronic databases.
• Appropriate and well-established decisions should be made using the data collected. It is practically
impossible to make sense of datasets containing more than a handful of data points without the help of
computer programs.
• To be certain of the insights that the collected data provides and to make further decisions, data mining is
performed where we go through distinctive analysis processes.
• Exploratory data analysis is key, and usually the first exercise in data mining. It allows us to visualize data to
understand it as well as to create hypotheses for further analysis.
• The exploratory analysis centers around creating a synopsis of data or insights for the next steps in a data
mining project. EDA actually reveals ground truth about the content without making any underlying
assumptions.
• This is the fact that data scientists use this process to actually understand what type of modeling and
hypotheses can be created.
• Key components of exploratory data analysis include summarizing data, statistical analysis, and visualization of
data.

Steps in EDA
• Having understood what EDA is, and its significance, let's understand the various steps involved in data
analysis.
• Basically, it involves four different steps. Let's go through each of them to get a brief understanding of
each step:
• Problem definition: Before trying to extract useful insight from the data, it is essential to define the
business problem to be solved. The problem definition works as the driving force for a data analysis plan
execution. The main tasks involved in problem definition are defining the main objective of the analysis,
defining the main deliverables, outlining the main roles and responsibilities, obtaining the current status
of the data, defining the timetable, and performing cost/benefit analysis. Based on such a problem
definition, an execution plan can be created.
• Data preparation: This step involves methods for preparing the dataset before actual analysis. In this
step, we define the sources of data, define data schemas and tables, understand the main
characteristics of the data, clean the dataset, delete non-relevant datasets, transform the data, and
divide the data into required chunks for analysis.

Steps in EDA..
• Data analysis: This is one of the most crucial steps that deals with descriptive statistics and analysis of
the data. The main tasks involve summarizing the data, finding the hidden correlation and relationships
among the data, developing predictive models, evaluating the models, and calculating the accuracies.
Some of the techniques used for data summarization are summary tables, graphs, descriptive statistics,
inferential statistics, correlation statistics, searching, grouping, and mathematical models.
• Development and representation of the results: This step involves presenting the dataset to the target
audience in the form of graphs, summary tables, maps, and diagrams. This is also an essential step as
the result analyzed from the dataset should be interpretable by the business stakeholders, which is one
of the major goals of EDA. Most of the graphical analysis techniques include scattering plots, character
plots, histograms, box plots, residual plots, mean plots, and others.

Making sense of data
• It is crucial to identify the type of data under analysis. In this section, we are going to learn about
different types of data that you can encounter during analysis.
• Different disciplines store different kinds of data for different purposes. For example, medical
researchers store patients' data, universities store students' and teachers' data, and real estate
industries storehouse and building datasets.
• A dataset contains many observations about a particular object. For instance, a dataset about patients
in a hospital can contain many observations.
• A patient can be described by a patient identifier (ID), name, address, weight, date of birth, address,
email, and gender. Each of these features that describes a patient is a variable. Each observation can
have a specific value for each of these variables. For example, a patient can have the following:

Making sense of data …
These datasets are stored in hospitals and are presented for analysis. Most of this data is stored
in some sort of database management system in tables/schema. An example of a table for
storing patient information is shown here:

• To summarize the preceding table, there are four observations (001, 002, 003, 004, 005). Each observation
describes variables (PatientID, name, address, dob, email, gender, and weight). Most of the dataset broadly
falls into two groups—numerical data and categorical data.
• Numerical data
This data has a sense of measurement involved in it; for example, a person's age, height, weight, blood
pressure, heart rate, temperature, number of teeth, number of bones, and the number of family members. This
data is often referred to as quantitative data in statistics. The numerical dataset can be either discrete or
continuous types.
• Discrete data
This is data that is countable and its values can be listed out. For example, if we flip a coin, the number
of heads in 200 coin flips can take values from 0 to 200 (finite) cases. A variable that represents a discrete
dataset is referred to as a discrete variable. The discrete variable takes a fixed number of distinct values. For
example, the Country variable can have values such as Nepal, India, Norway, and Japan. It is fixed. The Rank
variable of a student in a classroom can take values from 1, 2, 3, 4, 5, and so on.
• Continuous data
A variable that can have an infinite number of numerical values within a specific range is classified as
continuous data. A variable describing continuous data is a continuous variable. For example, what is the
temperature of your city today? Can we be finite? Similarly, the weight variable in the previous section is a
continuous variable. We are going to use a car dataset in Chapter 5, Descriptive Statistics, to perform EDA.

• Categorical data
This type of data represents the characteristics of an object; for example, gender, marital status, type of address, or categories of the
movies. This data is often referred to as qualitative datasets in statistics.
To understand clearly, here are some of the most common types of categorical data you can find in data:
◦ Gender (Male, Female, Other, or Unknown)
◦ Marital Status (Annulled, Divorced, Interlocutory, Legally Separated, Married, Polygamous, Never Married, Domestic Partner, Unmarried,
Widowed, or Unknown)
◦ Movie genres (Action, Adventure, Comedy, Crime, Drama, Fantasy, Historical, Horror, Mystery, Philosophical, Political, Romance, Saga, Satire,
Science Fiction, Social, Thriller, Urban, or Western
◦ Blood type (A, B, AB, or O)
◦ Types of drugs (Stimulants, Depressants, Hallucinogens, Dissociatives, Opioids, Inhalants, or Cannabis)
A variable describing categorical data is referred to as a categorical variable. These types of variables can have one of a limited number of
values. It is easier for computer science students to understand categorical values as enumerated types or enumerations of variables. There are
different types of categorical variables:
◦ A binary categorical variable can take exactly two values and is also referred to as a dichotomous variable. For example, when you create an
experiment, the result is either success or failure. Hence, results can be understood as a binary categorical variable.
◦ Polytomous variables are categorical variables that can take more than two possible values. For example, marital status can have several
values, such as annulled, divorced, interlocutory, legally separated, married, polygamous, never married, domestic partners, unmarried,
widowed, domestic partner, and unknown. Since marital status can take more than two possible values, it is a polytomous variable.

Comparing EDA with classical and
Bayesian analysis
There are several approaches to data analysis. The most popular ones that are relevant to this book
are the following:
• Classical data analysis:
For the classical data analysis approach, the problem definition and data collection step are followed by
model development, which is followed by analysis and result communication.
• Exploratory data analysis approach:
For the EDA approach, it follows the same approach as classical data analysis except the model
imposition and the data analysis steps are swapped. The main focus is on the data, its structure, outliers,
models, and visualizations. Generally, in EDA, we do not impose any deterministic or probabilistic models on the
data.
• Bayesian data analysis approach:
The Bayesian approach incorporates prior probability distribution knowledge into the analysis steps as
shown in the following diagram. Well, simply put, prior probability distribution of any quantity expresses the
belief about that particular quantity before considering some evidence. Are you still lost with the term prior
probability distribution? Andrew Gelman has a very descriptive paper about prior probability distribution. The
following diagram shows three different approaches for data analysis illustrating the difference in their
execution steps:

Comparing EDA with classical and
Bayesian analysis…

Software tools
• Python: This is an open source programming language widely used in data analysis, data mining, and
data science
• R programming language: R is an open source programming language that is widely utilized in statistical
computation and graphical data analysis
• Tableau: Tableau is a business intelligence and data visualization tool to get insights from data, find
hidden trends, and make business decisions.
• Weka: This is an open source data mining package that involves several EDA tools and algorithms
• KNIME: This is an open source tool for data analysis and is based on Eclipse

Visual Aids for EDA
• As data scientists, two important goals in our work would be to
• extract knowledge from the data
• present the data to stakeholders
• Presenting results to stakeholders is very complex in the sense that our audience may not have enough
technical know-how to understand programming jargon and other technicalities.
• Hence, visual aids are very useful tools
• Line chart, Bar chart, Scatter plot, Area plot, and stacked plot, Pie chart, Table chart, Polar chart,
Histogram.

Line chart
• A line chart is used to illustrate the relationship between two or more continuous variables.
• A line chart is a type of chart that provides a visual representation of data in the form of points that are
connected in a straight line.
• This is the most basic type of chart used in finance, and it typically only depicts a security's closing
prices over time.
• Line charts can be used for any timeframe, but they most often use day-to-day price changes
• When and how to use Line Charts for Visual Analysis?
◦ Data that is measured in a continuous progression works well in a line chart format. If your organization wants to track the behavior
of data over a given period, line charts can be of great help.
◦ Your audience will see changes in the data, plotted out with a line connecting each data point as they changed over time.

Line chart…..
• This line chart follows the number of borrowers
measured every month over the course of
several years.
• The x-axis represents the Funded date field
• The y-axis represents the No of borrowers
• The first and last points are labeled
• Labels for each data point are not displayed to
reduce clutter
• Axis and light grid lines are visible to the viewer
to help in identifying locations of data points

Bar charts
• Bars can be drawn horizontally or vertically to represent categorical variables
• Bar charts are frequently used to distinguish objects between distinct collections in order to track
variations over time.
• A bar chart or bar graph is a chart that represents categorical data with rectangular bars with heights
proportional to the values that they represent. Here one axis of the chart plots categories and the other
axis represents the value scale. The bars are of equal width which allows for instant comparison of data.

Bar chart…
• A pharmacy in Norway keeps track of the
amount of Zoloft sold every month. Zoloft is a
medicine prescribed to patients suffering from
depression
• The x-axis represents the month
• The Y- axis represents the quantity of medicine

Scatter plot
• Scatter plots are also called scatter graphs, scatter charts, scatter grams, and scatter diagrams.
• Scatter plots are used to plot data points on a horizontal and a vertical axis in the attempt to show how
much one variable is affected by another.
• When should we use a scatter plot? Scatter plots can be constructed in the following two situations:
• When one continuous variable is dependent on another variable, which is under the control of the observer
• When both continuous variables are independent
• Some examples in which scatter plots are suitable are as follows:
• Research studies have successfully established that the number of hours of sleep required by a person depends on the age of the
person.
• The average income for adults is based on the number of years of education.

Scatter plot..
• Research studies have successfully established
that the number of hours of sleep required by a
person depends on the age of the person
• The x-axis represents the age of persons in years
• The Y- axis represents the total hours of sleep
required

Pie chart
• A pie chart, sometimes called a circle chart, is a way of summarizing a set of nominal data or displaying
the different values of a given variable (e.g. percentage distribution).
• This type of chart is a circle divided into a series of segments.
• Each segment represents a particular category. The area of each segment is the same proportion of a
circle as the category is of the total data set

Pie chart..
• For example, take a pie chart representing the
sales a retail store makes in a month. The slices
represent the sales of toys, home decor,
furniture, and electronics.
• The below pie chart represents sales data,
where 43 percent of sales were from furniture,
28 percent from electronics, 15 percent from
toys, and 14 percent from home decor.
• These figures add up to 100 percent, as should
always be the case with pie graphs

Data Transformation
• Data transformation is a set of techniques used to convert data from one format or structure to another
format or structure.
• The following are some examples of transformation activities:
• Data deduplication involves the identification of duplicates and their removal. Key restructuring involves
transforming any keys with built-in meanings to the generic keys.
• Data cleansing involves extracting words and deleting out-of-date, inaccurate, and incomplete
information from the source language without extracting the meaning or information to enhance the
accuracy of the source data.
• Data validation is a process of formulating rules or algorithms that help in validating different types of
data against some known issues.
• Format revisioning involves converting from one format to another.
• Data derivation consists of creating a set of rules to generate more information from the data source.

Data transformation..
• Data aggregation involves searching, extracting, summarizing, and preserving important information in
different types of reporting systems.
• Data integration involves converting different data types and merging them into a common structure or
schema.
• Data filtering involves identifying information relevant to any particular user.
• Data joining involves establishing a relationship between two or more tables.
The main reason for transforming the data is to get a better representation such that the transformed
data is compatible with other data. In addition to this, interoperability in a system can be achieved by
following a common data structure and format

Merging database-style dataframes
• Many beginner developers get confused when working with pandas dataframes, especially regarding
when to use append, concat, merge, or join
• Let's assume that you are working at a university as a professor teaching a Software Engineering course
and an Introduction to Machine Learning course, and there are enough students to split into two
classes. The examination for each class was done in two separate buildings and graded by two different
professors. They gave you two different dataframes. In the first example, let's only consider one
subject— the Software Engineering course

Merging..
• In the preceding dataset, the first column contains information about student identifiers and the second
column contains their respective scores in any subject. The structure of the dataframes is the same in
both cases. In this case, we would need to concatenate them.
• The output of the preceding code is a single dataframe combining both of the tables. These tables
would be merged into a single one as shown in the following screenshot:

Reshaping and pivoting
• During EDA, we often need to rearrange data in a dataframe in some consistent manner.
• This can be done with hierarchical indexing using two actions:
• Stacking: Stack rotates from any particular column in the data to the rows.
• Unstacking: Unstack rotates from the rows into the column.
• The output of the preceding snippet is as follows:
• Now, using the stack() method on the preceding dframe1, we can pivot the columns into rows to
produce a series

Transformation techniques
• let's dive more into how we can perform ther types of data transformations including cleaning, filtering,
deduplication, and others.
• 1. Performing data deduplication
• It is very likely that your dataframe contains duplicate rows. Removing them is essential to enhance the
quality of the dataset.
• You can clearly see from the following screenshot that in both columns, there are some duplicate
entries:
• The rows that say True are the ones that contain duplicated data.

Transformation techniques..
2. Replacing values Often, it is essential to find and replace some values inside a dataframe. This
can be done with the following steps:

3. Handling missing data
• Whenever there are missing values, a NaN value is used, which indicates that there is no value specified for that
particular index.
• There could be several reasons why a value could be NaN:
• It can happen when data is retrieved from an external source and there are some incomplete values in the dataset.
• It can also happen when we join two different datasets and some values are not matched.
• Missing values due to data collection errors.
• When the shape of data changes, there are new additional rows or columns that are not determined.
• Reindexing of data can result in incomplete data.
• NaN values either can be changed to true using isnull(), or false using notnull()

4. Dropping missing values
• One of the ways to handle missing values is to simply remove them from our dataset.
• Dropping by rows - We can also drop rows that have NaN values.
• Dropping by column- We can also drop columns that have NaN values
5. Filling missing values
• We can fill the missing values with some values
• Backward and forward filling NaN values can be filled based on the last known values
6. Outlier detection
• Outliers are data points that diverge from other observations for several reasons. During the EDA phase, one of our
common tasks is to detect and filter these outliers.
• The main reason for this detection and filtering of outliers is that the presence of such outliers can cause serious issues in
statistical analysis.

Grouping Datasets
• During data analysis, it is often essential to cluster or group data together based on certain criteria.
• For example, an e-commerce store might want to group all the sales that were done during the
Christmas period or the orders that were received on Black Friday.
• The pandas groupby function is one of the most efficient and time-saving features for doing this.
Groupby provides functionalities that allow us to split-apply-combine throughout the dataframe; that is,
this function can be used for splitting, applying, and combining dataframes.
• The pandas groupby method performs two essential functions:
• It splits the data into groups based on some criteria.
• It applies a function to each group independently
• Let's take a look at a dataset of automobiles that enlists the different features and attributes of cars,
such as symbol ling, normalized-losses, make, aspiration, body-style, drive-wheels, engine-location, and
many others.

Grouping Datasets…
lets us group this dataset on the basis of the body-style column:

Data aggregation
• Aggregation is the process of implementing any mathematical operation on a dataset or a subset of it
• Some of the most frequently used aggregations are as follows:
• sum: Returns the sum of the values for the requested axis
• min: Returns the minimum of the values for the requested axis
• max: Returns the maximum of the values for the requested axis
• Since aggregation only works with numeric type columns, let's take some of the numeric columns from
the dataset and apply some aggregation functions to them:
• Next, let's apply a single aggregation to get the mean of the columns
• We can aggregate more than one function together. For example, we can find the sum and the
minimum of all the columns at once

Data aggregation …
Sum for all column
length 0.837102
width 0.915126
height 53.766667
curb-weight
2555.666667
price 13207.129353

EDA-Unit 1.pdf

More Related Content

What's hot (20)

Similar to EDA-Unit 1.pdf (20)

More from Nirmalavenkatachalam (9)

Recently uploaded (20)

EDA-Unit 1.pdf