A Data Analysis Using Python

By: Emmanuel K. Musyoki

Introduction

This notebook aims at cleaning, analyzing, and visualizing a CSV World Bank data on GDP growth in different countries and regions of the world FROM 2010 TO 2021

Section 1: Exploratory Data Analysis

• Removing unwanted columns
• Checking and Handling Missing Data (Null Values)
• Checking any anomalies and plotting them
• Visualizing the data
• Conclusion

In [86]:

#importing modules
import pandas as pd
import numpy as np
import matplotlib as mpl
import matplotlib.pyplot as plt
import seaborn as sns

In [87]:

# Load the dataset
Gdp_data = pd.read_csv (r"C:\Users\USER\Desktop\PYCHARM\csv _excel_ files\economic_indicator2.csv")

In [88]:

Gdp_data.head()

Out[88]:

A Screenshot of the Dataset head after running the code above

In [89]:

Gdp_data.describe()

Out[89]:

8 rows × 61 columns

• We Want data from 2010 to 2021 only,
• delete the rest of the year's columns and the indicator code column

In [90]:

Gdp_data.drop(Gdp_data.iloc[:,3:53],axis = 1,inplace = True)

Now I remain with the data I want to work on, let's see the information after deleting unwanted columns

In [91]:

Gdp_data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 266 entries, 0 to 265
Data columns (total 15 columns):
 #   Column          Non-Null Count  Dtype  
---  ------          --------------  -----  
 0   Country Name    266 non-null    object 
 1   Country Code    266 non-null    object 
 2   Indicator Name  266 non-null    object 
 3   2010            254 non-null    float64
 4   2011            254 non-null    float64
 5   2012            254 non-null    float64
 6   2013            254 non-null    float64
 7   2014            256 non-null    float64
 8   2015            255 non-null    float64
 9   2016            254 non-null    float64
 10  2017            254 non-null    float64
 11  2018            254 non-null    float64
 12  2019            252 non-null    float64
 13  2020            249 non-null    float64
 14  2021            244 non-null    float64
dtypes: float64(12), object(3)
memory usage: 31.3+ KB

HANDLING MISSING DATA

1. Checking Null Values

Total Number of Null Rows in every column

In [92]:

Gdp_data.isnull().sum()

Out[92]:

Country Name       0
Country Code       0
Indicator Name     0
2010              12
2011              12
2012              12
2013              12
2014              10
2015              11
2016              12
2017              12
2018              12
2019              14
2020              17
2021              22
dtype: int64

• Filling Null Values With the mean of individual column
• Naming the new dataset without Null values Gdp_data1

In [93]:

Gdp_data1= Gdp_data.fillna(value =Gdp_data['2010'].mean())

In [94]:

Gdp_data1.isnull().sum()

Out[94]:

Country Name      0
Country Code      0
Indicator Name    0
2010              0
2011              0
2012              0
2013              0
2014              0
2015              0
2016              0
2017              0
2018              0
2019              0
2020              0
2021              0
dtype: int64

There is No Null Value Now As seen above, we have replaced them with the mean of the column

In [95]:

Gdp_data1.head()

• Rounding Off the values in every column to one decimal place

In [96]:

Gdp_data1.round(1)

Out[96]:

Cleaned Data

266 rows × 15 columns

Cleaned Data

START VISUALIZATION

• You can visualize the GDP growth for every year or for each country in different years
• You can also compare the GDP growth for two years.

In [97]:

import matplotlib.pyplot as plt
import seaborn as sns
import warnings
warnings.filterwarnings('ignore')

In [145]:

plt.figure(figsize=(10,6))
sns.distplot(Gdp_data1['2010'], kde=False, hist=True)
plt.title('GDP percentage growth', size=16)
plt.ylabel('Country count');

A bar graph

From The above Graph we can see that;

• majority of countries' GDP growth lies between 2% and 7%.
• there are a few outliers

Identifying Outliers

• Selecting Columns to plot

In [99]:

To_plot = Gdp_data1.drop(columns=['Country Code','Indicator Name' ]).select_dtypes(include=np.number)

Subplots

In [100]:

To_plot.plot(subplots=True, layout=(4,4), kind='box', figsize=(12,14), patch_artist=True)
plt.subplots_adjust(wspace=0.5)

Out [100]:

Box plots

• majority of countries' GDP growth is between 0 to 6%
• The year 2020 has negative GDP growth for all countries

Statistical Analysis

• Step 1 Filtering the Data to remain with 10 countries only
• I will call it Gdp_data2

In [102]: Dropping columns to remain with 2020 and 2021 only and 10 countries

Gdp_data2= Gdp_data1.drop(Gdp_data1.iloc[:,3:13],axis = 1,inplace =False)
Gdp_data2.info()

Gdp_data2.drop(Gdp_data2.index[11:267], inplace=True)

Plot the GDP growth rate in 2020

In [142]:

plt.figure(figsize=(5,4))
Gdp_data2.groupby('Country Name')['2020'].sum().sort_values(ascending=False).plot(kind='bar')
plt.title('2020 GDP growth rate', size=16)
plt.ylabel('percentage rate');

A screenshot of output in 142

Plot the GDP growth rate in 2021

In [143]:

plt.figure(figsize=(5,4))
Gdp_data2.groupby('Country Name')['2021'].sum().sort_values(ascending=False).plot(kind='bar')
plt.title('2021 GDP Growth Rate', size=16)
plt.ylabel('percentage rate');

Observation for the two Years

• In 2020, The economy of every country was a negative GDP
• The covid 19 pandemic caused this.
• Aruba had the highest GDP decline followed by Andora in 2020 while in 2021 Aruba had the highest GDP growth followed by Argentina
• Africa Western and Central were least affected by the pandemic in 2020 since the GDP dropped to -0.1 only.