Exploratory Data Analysis using Python

Another post starts with you beautiful people!
In my previous posts and pages we have learnt basics and advanced topics of Python required in Data Science.
Now it's time to do EDA, sounds interesting!

Exploratory Data Analysis (EDA) is a crucial step of the data analytics process.
It involves exploring the data and identifying important features about the data as well as asking interesting questions from the data by using statistical and visualization tools studied in earlier classes such as descriptive statistics and basic plotting.

In this post we will use the dataset about TB data on countries and their territories.
Specifically, we would using data files for TB Deaths, spread of TB, and number of new cases of TB to answer some important questions.

Since we are going to perform some Exploratory Data Analysis in our TB dataset, these are the questions we want to answer:

• Which are the countries with the highest and infectious TB incidence?
• What is the general world tendency in the period from 1990 to 2007?
• What countries don't follow that tendency?
• What other facts about the disease do we know that we can check with our data?

First set the local path where you want to put files for example I am using-

Second, import required libraries-

Next, we will get our dataset from the internet resource and save those in our local disk-

For more details about urllib.request library please visit here- tell me more about urllib.request

After the above step the dataset will be saved in your local path as csv files and we are ready to use these datasets.

Now we will read the csv files and do some beautification -

After this let's explore few data-

or

Result:-

If you want to check percentage change in existing cases over the years-

Let us look at curious case of Spain. What do you infer?

Let us go ahead and do some plotting-

How about box-plots-

Which country has the highest number of existing and new TB cases?

Result:

What about world trends?

Result:

What about specific countries?

Result:

Let us think about outlier countries-

Proportions of countries as outlier-

Filter the data frame:-

Result:

What do you infer from above dataset? Can we somehow combine all of that information?

Compare this with rest of world:-

Result:

What about percentage change?

Result:-

Let's see TB cases in China-

Hope you enjoyed today' learning.

If you are reading and practicing these learnings then no doubt you are a future data scientist.

Last but not the least as a data scientist- ASK THE RIGHT QUESTION !