Analyzing and Visualizing Real-World Data

Analyzing and Visualizing Real-World Data

0%

1. Preprocessing Data: Part I

First, let's get acquainted with the data we will be working with. Throughout this section, we will perform the initial set of actions to prepare the data for further analysis.

Loading the Data

Columns Overview

Removing Characters: Method 1

Removing Characters: Method 2

Checking Columns

Replacing Characters

2. Preprocessing Data: Part II

In the second section, we will continue with the preparatory tasks started in the first section and ensure the data is ready for further exploration.

Analyzing Columns

Removing and Replacing

Processing Dates

Wrong Convertation?

Proper Converting

3. Analyzing Data

After preparing the data, we will move on to some exploratory analysis. Throughout this section, we will answer several questions and uncover interesting insights.

Superficial Analysis

The Most Profitable Store

How Periodic the Data is?

The Most Profitable Periods

The Best Combination

Holidays Effect

What Holidays are we Considering?

The Most Profitable Holiday

4. Visualizing Data

In the last section, we will use visualization tools to either confirm or reject the hypotheses we made in the previous section. We will also create visualizations of key indicators for the entire dataset.

First Look at the Data

Comparing Dynamics

Finding Relation

Does Temperature Affect?

Or is it Holiday Effect?

Comparing Shops

Adding Holidays to Comparison

Columns Overview

We can see that only two types of data are presented in the dataframe: object and int64. Columns that have an object data type contain string objects, which makes aggregation impossible for them. Among these columns are Weekly_Sales, Temperature, Fuel_Price, and others. It is obvious that all the mentioned columns must be numerical. We might be interested in comparing the revenue for different dates, but with object data, it's impossible.

Let's solve problems step by step. First, let's remind ourselves what our data looks like by outputting a single row.

Pay close attention to the values and try to find out why most of the columns are considered object columns.

Everything was clear?

Section 1. Chapter 3

Course Content

Analyzing and Visualizing Real-World Data

Analyzing and Visualizing Real-World Data

Columns Overview

We can see that only two types of data are presented in the dataframe: object and int64. Columns that have an object data type contain string objects, which makes aggregation impossible for them. Among these columns are Weekly_Sales, Temperature, Fuel_Price, and others. It is obvious that all the mentioned columns must be numerical. We might be interested in comparing the revenue for different dates, but with object data, it's impossible.

Let's solve problems step by step. First, let's remind ourselves what our data looks like by outputting a single row.

Pay close attention to the values and try to find out why most of the columns are considered object columns.

Everything was clear?

Section 1. Chapter 3

some-alt