Learners will likely run into errors before any episode dedicated to errors is covered. Also, having an episode about errors is helpful but can be difficult to use or follow if necessary info is separate from the episode being taught.

Is there some way to include information about troubleshooting common errors within the lesson? Perhaps as a reference or callout that would support independent learning in situations where someone is using the content outside of a workshop context?

The example dataset is large. Need to identify the size of a useful subset and do the pre-processing to create one.

There are no NaNs in the data subset. Find and include some raw data with missing values, or alternatively add some random NAs.

Jupyter Notebooks are not intuitive and can increase the cognitive load of novice learners because:

• the service launches as a server, unlike a typical desktop application
• the file browser is confusing
• the user 'home' and r/w permissions are platform dependent

Also, unless learners have an immediate need to use Jupyter Notebooks after the lesson/workshop, they may forget how to use the service and not have easy access to workshop content, examples, etc.

Time series data have unique features and relevant concepts with regard to descriptive statistics - rolling means, etc.

It would be worthwhile to revise the whole lesson to incorporate these concepts, or alternatively keep the existing introductory content and add a lesson after pandas basic concepts that introduces time series data concepts.

Provide a "python fundamentals" episode similar to what's already in SWC Programming with Python.

Several topics are introduced in episode 2 that need more detailed development - lists, loops, slicing lists, etc.