Data Analysis Bootcamp™ 21 Real World Case Studies

Kick off your data science journey by mastering Python basics: variables, types, printing, type checks, and simple arithmetic and string concatenation in Google Colab.

Master how to define and use Python functions with def, return area and volume calculations for circle and cylinder, and apply default and named parameters for clearer code.

Define a simple Python class using a book example with __init__ and attributes like title and author. Print or update the instance to see data stored and reuse enabled.

Learn to work with pandas series, from importing pandas as pd to converting lists into series and applying max, min, and sort. Understand index versus values and inplace updates.

Explore pandas dataframes by loading CSVs, saving to CSV or Excel, and converting data types; index, slice, and filter rows and columns; apply describe, mean, median, mode, and value_counts.

Apply data cleaning with pandas: load a dataset in Colab using iso-8859-1, alter columns and rows, rename to English, identify states, and fix missing or non-numeric values with string checks.

Learn practical pandas data cleaning: strip and replace text, handle missing values with fill (backfill and forward fill), convert to floats, drop columns or rows, reset indices, and map strings.

Master feature engineering in pandas by building a new family size feature via apply and lambda on the Titanic dataset.

Learn to concatenate, merge, and join dataframes to stack, align columns, and handle index differences. Apply concat, append, and left, right, inner, and full outer joins to combine datasets.

Create an hourly time series in Pandas from a date range and build a dataframe with a date index. Resample to daily means and explore date parsing and formatting options.

Master advanced Pandas operations with map, zip, and apply. Learn how to apply functions to lists, pair sequences, and filter results efficiently.

Explore choropleth maps with Plotly, building US state and county visualizations from FIPS data, using color scales and legends, and extending to world life expectancy, GDP, and other maps.

Learn how descriptive statistics summarize data to reveal patterns, rely on simple visualizations, and apply exploratory data analysis to understand data structure.

Explore exploratory data analysis (EDA) with visualizations like histograms, violin plots, box plots, and scatter plots in Python, pandas, and seaborn on the wine quality data set.

Explore advanced exploratory data analysis visualizations using Seaborn, including nested bar plots, density plots, joint distributions, factor plots, and CDF-based quantiles across datasets like Titanic.

Learn how variance and standard deviation measure data spread, apply Bessel's correction, and compare with range and mean absolute distance, using Python and pandas on a wine dataset.

Classify data as quantitative or qualitative, and distinguish nominal, ordinal, interval, and ratio scales—discrete and continuous variables—using real-world examples and frequency distributions.

Explore frequency distributions and histograms to summarize data, choose appropriate bins, and interpret the distribution of values such as wine alcohol percentages for informed decisions.

Explore the shapes of frequency distributions, including left and right skew, normal and uniform distributions, with real-world examples like human heights and weight to understand how tails reveal skewness.

Explore covariance and correlation as measures of how two variables vary together. Learn normalization, interpret correlation from -1 to 1, and visualize relationships with pandas, seaborn, and correlation matrices.

Choose random, representative samples to estimate population means and reduce sampling error. Stratify by wine type and compare red and white means to show how larger sample sizes improve accuracy.

Explore the normal distribution and the central limit theorem, showing how sample means form a normal distribution from any population and how z scores relate.

Learn how correlations can mislead by showing why divorce rates and margarine use do not imply causation, and examine spurious links before the normal distribution and central limit theorem.