Delve into Python data analysis and visualization using NumPy, SciPy, Pandas, and Seaborn. Gain insights into reading, cleaning, and visualizing data to generate actionable insights.

mlxtend.tar.gz

visualization

This course assumes no prior experience and can teach anyone how to use Python to analyze and visualize data. Stop using Excel and start harnessing the power of Python.

Data analysis is one of the hottest careers of the 21st century. As an analyst, your goal is to peel back layers of data in order to answer questions of interest; that is the power of analytics. It allows you to take raw data and create meaningful, actionable insights.

In this course, you’ll learn how to use Python, NumPy, SciPy, Pandas, and Seaborn to perform data analysis and visualization. You’ll explore the four crucial steps for any data analysis project: reading, describing, cleaning, and visualizing data. In each step, you will work with the most common and popular tools that data analysts use every day. By the end of the course, you will be able to confidently extract knowledge and answers from data.

Python Data Analysis and Visualization

# What is missing data?

So, you have read in your data only to discover that some values are missing! What do you do?

First, you should try to understand why your data points are missing. Are they missing at random or not? Data that is missing at random could be removed with a large enough dataset. If your data is missing for a reason, perhaps missing means zero? Or is it a strong signal of a sensor malfunction? These non-random missing values should be fixed or leveraged. For example, set them to zero if they should be zero.

Second, you should consider how much data you have relative to how many rows have missing data. If you have 1 million data points with 10 points missing at random, you are probably okay to drop those rows. On the other hand, if you have 300 data points with 100 are missing, then you should probably be hesitant about removing 1/3 of your data. Assuming you can't find an underlying reason for the missing data, how do you handle them?

# Filling missing data

There are a few ways to fill in missing values.

## Using a statistical value

Fill in a column's missing values with a **statistical value** such as the mean, median, or mode. This has the advantage of being simple and easy to do but also introduces a smaller variance within your column than otherwise would not be present. High variance within a column is usually beneficial (assuming it is real) to a machine learning model. Also, it might not make a lot of sense. For example, imagine a data set of heights and weights with missing weights. You then fill all the missing weights with the average regardless of height. We know that height and weight are correlated, but we are not taking advantage of that knowledge.

Let's look at a simple example of our first idea of filling in missing values with a statistical value.

This lesson focuses on missing data by explaining some techniques to clean such data from our dataset.

What is Analytics

Python Basics for Analytics

Reading Data

Describing Data

Cleaning Data

Visualizing Data

Missing Data

What is missing data? #