Data Analysis Bootcamp 1 - Data Analysis Intro
In this Data Analysis Bootcamp class, we will focus on honing your data-driven decision-making skills by investigating variable relationships, uncovering correlations, and enabling you to evaluate alternatives and assess risks effectively. We will also delve into identifying external opportunities and problems, such as market trends and customer preferences, while addressing internal process inefficiencies to enhance organizational performance.
Data Analysis Bootcamp 3 - Pandas Plotting Hell Week 1
In this data analysis bootcamp class using Python, we will harness the user-friendly plotting tools built directly into Pandas, enabling us to delve into exploratory data analysis. Our focus will begin with univariate data exploration, employing tools such as histograms, area plots, and boxplots to gain insights into the distribution and characteristics of individual variables. Moving on, we'll venture into the bivariate realm using scatter matrices to uncover relationships between pairs of variables. Additionally, we'll tap into Pandas' high-level statistical plots, including autocorrelation and Andrews curves, to deepen our understanding of data patterns. Lastly, we'll emphasize the ease of highlighting and formatting our plots and dataframes in Pandas, enhancing our ability to effectively communicate analytical results.
Data Analysis Bootcamp 5 - Bivariate Analysis
In our fifth data analysis bootcamp, we explore bivariate analysis, a vital aspect of data science focused on understanding relationships between two variables. This exploration equips us with tools to uncover intricate data connections, leading to valuable insights and informed decision-making. By grasping these relationships, we can predict trends and mitigate risks, crucial in our data-driven world. Bivariate analysis goes beyond identifying relationships; it quantifies their strength and direction, enhancing our ability to make data-based decisions with statistical rigor. Join us on this journey to unveil hidden data stories and harness their potential for informed decision-making.
Data Analysis Bootcamp 7 - Wrangling, Cleaning, Treating
In our 7th data analysis class, we focus on creating and joining datasets, including operations like joining and concatenating data, which are crucial for consolidating information from various sources. Data cleaning is another significant aspect, addressing issues such as spelling corrections and handling outliers, both vital for maintaining data accuracy. We also delve into the effects of treating outliers, equipping students with the knowledge needed for robust data analysis.
In addition to data integration, we place a strong emphasis on data cleaning in this class. This entails rectifying issues such as misspellings, missing values, and handling outliers. Correcting spelling errors is crucial to ensure data consistency and accuracy. Handling outliers, on the other hand, is essential for maintaining the integrity of our analyses. We explore techniques for detecting and addressing outliers, which can significantly impact the outcomes of our data analysis. Understanding the effects of treating outliers and the various methodologies to do so is a pivotal component of this class, ensuring that we are well-equipped to perform robust data analysis.
Data Analysis Bootcamp 9 - Hacker Statistics
In our ninth class of the data analysis bootcamp, we delved into a crucial concept closely linked to our initial discussion on sampling. We emphasized the profound impact that sampling has on the comprehension of summary statistics. When we take a sample from a larger dataset, we introduce an inherent element of randomness that cannot be precisely measured but exerts a tangible influence on the insights we can derive from our data.
To better appreciate this randomness and its implications, we introduced the concept of Hacker Statistics, specifically focusing on the Bootstrap resampling technique. Hacker Statistics, using Bootstrapping, provides us with a powerful tool to understand and quantify the uncertainty associated with our data. Through the application of Hacker Statistics, we can simulate hypothesis testing scenarios and gain valuable insights into the reliability of our statistical inferences. This newfound capability enables us to visualize and interpret our data in a more comprehensive and robust manner, ultimately enhancing our data analysis skills.
Data Analysis Bootcamp 2 - Understanding Distributions
In this data analysis class, we will explore the essential principles of univariate analysis, which involves examining individual variables in isolation to gain insights into their distributions and characteristics. Understanding these data distributions is crucial for determining central tendencies, variabilities, and patterns, enabling us to make informed decisions, detect outliers, and select suitable statistical tests or machine learning techniques. Univariate analysis serves as the foundation for more complex multivariate analyses and statistical modeling approaches.
Data Analysis Bootcamp 4 Seaborn Univariate Hell Week 2
In this data analysis bootcamp class number 4, we will do a comprehensive walkthrough of Seaborn's powerful univariate analysis tools. Our primary objective is not only to understand what these tools can do, but also to understand the nuances of when, where, and how to effectively utilize them. We will cover essential Seaborn plots such as histplot, kdeplot, swarm plot, countplot, as well as figure level plots like displot and catplot. This class is particularly well-suited for those who are new to Seaborn and the world of data analysis in Python. So, let's dive in and unlock the potential of Seaborn for insightful data analysis!
Data Analysis Bootcamp 6 - Hell Week 3 - Seaborn Bivariate
In our sixth data analysis bootcamp class, we embark on a journey through the intricacies of data visualization. Beginning with fundamental plots like scatter plots and regression plots in Seaborn, we establish a solid foundation for effective data representation. As we advance, we explore more sophisticated visualization techniques, including the essential jointplot and heatmap plot, which are pivotal in modern data analysis. Moving on, we delve into the visualization of multivariate data with PairPlot and PairGrid, expanding our capabilities in comprehending complex data relationships. To cap off this phase, we introduce the LMPlot, a figure-level linear model plot that allows us to gain deeper insights into data interactions. In closing, we engage in a thought-provoking discourse on the significance of diverging color palettes in bivariate analysis, enriching our understanding of the intricate world of data analysis.
Data Analysis Bootcamp 8 - Interactive Plotly
In the 8th class of our Python data analysis bootcamp, we move our focus on to the powerful plotting engine, Plotly. Although Plotly is used to build Dashboard's Plotly express allows for quick and easy one-off plots which serves perfectly for our data analysis needs.Â
These one-off interactive plots allow for high-level data analysis on the spot allowing us to go deeper and extract more insights with a single plot.  And go a further in our analysis without needing to go back to pandas to understand all sides of the patterns we notice.
By leveraging Plotly Express, we can construct visualizations that offer a contrast and complement to the capabilities of libraries like Pandas and Seaborn.
With plots like the Sunburstplot which allow us to uncover deeper understanding of our categories's variable and their interrelationships. The 3D Scatter plot allows for an unparalleled understanding of the relationships in our continuous variable in our data.
Furthermore, we will introduce plot formatting in Plotly with fig.update_layout method. This allows us to control margins and titles. To make our Plotly plots not only insightful but beautiful as well.