-
Created a
data
folder in your local project repository. -
Downloaded the two CSV files and place them in the data folder:
a. Gross Domestic Product (GDP) per capita http://data.un.org/Data.aspx?d=WDI&f=Indicator_Code%3aNY.GDP.PCAP.PP.KD
- rename the file to
gdp_percapita.csv
- open it with a text editor (not excel) and take a look
b. Percentage of Individuals using the Internet http://data.un.org/Data.aspx?d=ITU&f=ind1Code%3aI99H
- rename the file to
internet_use.csv
- open it with a text editor and take a look
- rename the file to
-
Launched a Jupyter Notebook.
- Used markdown cells to record answers to any questions asked in the exercise. On the menu bar, you can toggle the cell type from
Code
toMarkdown
.
-
Imported the required packages with their customary aliases as follows:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
-
Used the
%matplotlib inline
command so that your plots show in the notebook without having to callplt.show()
every time. -
Using the pandas
read_csv()
function, read the GDP dataset into your notebook as a DataFrame calledgdp_df
. Take a look at the first 6 rows. -
Repeated the above steps for the internet use dataset. Call this DataFrame
internet_df
. Looked at the first six rows. -
Looked at the shape of each dataframe - how many rows, how many columns.
-
Looked at the data types for the columns in each table.
-
Looked at the last 10 rows of each dataset in turn.
-
Dropped the 'value footnotes' data (column) from both datasets. Check that this worked as expected.
-
Changed the columns for the GDP Per Capita data frame to ‘Country’, ‘Year’, and ‘GDP_Per_Capita’.
-
Changed the columns for the Internet Users data frame to ‘Country’, ‘Year’, and ‘Internet_Users_Pct’.
-
Merged the two DataFrames to one. Merge all rows from each of the two DataFrames. Call the new DataFrame
gdp_and_internet_use
. -
Looked at the first five rows of your new data frame to confirm it merged correctly.
-
Looked at the last five rows to make sure the data is clean and as expected.
-
Subset the combined data frame to keep only the data for 2004, 2009, and 2014. Checked that this subset correctly.
-
Created three new data frames, one for 2004, one for 2009, and one for 2014. Gave them meaningful names aand answered the questions below.
-
Which country had the highest percentage of internet users in 2014? What was the percentage? (typing the first 3 letters of your DataFrame name and hitting the tab for auto-complete options).
-
Which country had the lowest percentage of internet users in 2014? What was the percentage?
-
Repeated the above steps for 2004 and 2009.
-
Which country had the highest gdp per capita in 2014? What was the gdp per capita?
-
Which country had the lowest gdp per capita in 2014? What was the gdp per capita?
-
Created some scatterplots:
a. 2004 Percent Using the Internet vs GDP Per Capita
b. 2009 Percent Using the Internet vs GDP Per Capita
c. 2014 Percent Using the Internet vs GDP Per Capita -
Deterime there differences across years? What do the plots tell you about any relationship between these two variables? Enter your observations as a markdown cell.
-
Looked at the distribution of gdp per capita values for 2014. Is it unimodal?
-
Looked at the distribution of Internet Use for 2014. Is it unimodal?
-
What are the top 5 countries in terms of internet use in 2014?
-
Created a data frame called top_5_internet from the combined data frame that has all three years for these 5 countries. You should have 15 rows. Check that this is true.
-
Created a seaborn FacetGrid to show the internet usage trend over time for these 5 countries (those with the highest reported internet use in 2014). Which country had the greatest growth between 2004 and 2014? Is there a plotting issue with Bermuda? Can you fix it?
-
Repeated the steps above to look at the trend for the 5 countries with the lowest 2014 internet usage. Which country has consistently had the least internet use?
-
Found the top 5 countries for 2014 in terms of GDP per capita; create a dataframe to look at 10-year trends in gdp per capita for those 5 countries. Use a seaborn facet grid for this.
-
Repeated this one more time to look at 10-year trend for the bottom 5 countries for 2014 in terms of GDP per capita.
-
Is there anything surprising or unusual in any of these plots? Searching on the internet, can you find any possible explanations for unusual findings?
- Download another data set from the UN data (http://data.un.org/Explorer.aspx) to merge with your data and explore.