Data Manipulation with Pandas Masterclass

Explore data manipulation with pandas in this masterclass, handling tabular data and core operations using Google Colab for hands-on labs, alongside Excel and SQL.

discover tabular data fundamentals and master pandas operations, then visualize insights with pandas and seaborn, followed by a hands-on practice session.

Explore tabular data with rows and columns from Excel, CSV, TSV, or databases, and how pandas imports these sources for a data point, features, labels, and categorical versus continuous data.

master data manipulation with pandas from the 1.0 milestone, using series and data frames, with input/output, selection, filtering, feature engineering, aggregation, pivot tables, joins, and plotting with matplotlib.

Examine data structures in Pandas, comparing the series—one-dimensional with an index—and the data frame—two-dimensional with rows and columns. Learn how selecting a column returns a series within a unified framework.

Learn how pandas reads data from multiple formats—Excel, Parquet, HTML, JSON, and databases like Google BigQuery and SQL—using read_csv, then exports to formats for sharing.

Learn Pandas data selection on a Titanic data frame: single versus double bracket outputs, condition filters with or logic, and row-column selection with loc and iloc.

Learn when to import numpy with pandas, since pandas loads numpy and data frames can be converted to numbers; use numpy functions like np.log for math not in pandas.

Explain zero-indexed range notation in pandas, with start inclusive and end exclusive, using nine to twenty five for rows and two to five for columns.

Master feature engineering in pandas by creating new columns and applying element-wise operations, including multiplying by a fixed number and vectorized division of columns.

Explore Pandas aggregations and summary statistics, including mean calculations, group by operations, multiple aggregations with agg, and counting with value_counts for passenger class analysis.

Sort values with sort_values by column and order, then pivot data between long and wide formats and build pivot tables with index, columns, and mean aggregation.

Learn to join data by concatenating frames with the same shape along axis zero or one, and merge two frames on a key using a left join.

Convert a date and time column to datetime objects with pandas to_datetime, yielding timestamps or timedeltas. Set the datetime index for time series plotting, as pandas recognizes datetime indices.

Discover how in-memory data storage and operations returning new data frames, as seen with joins, limit big data handling when memory runs short.

Explore pandas tools for quick data inspection with head, tail, info, and describe, plus get dummies for categorical encoding and robust null handling, rolling and exponentially weighted moving average analytics.

Explore data visualization in pandas, using plot, scatter, bar, histogram, area, and violin plots, with seaborn as a high-level, one-line alternative built on matplotlib.