Data science best practices with pandas (video tutorial)
The pandas library is a robust software for a number of phases of the information science workflow, together with information cleansing, visualization, and exploratory information evaluation. Nonetheless, the measurement and complexity of the pandas library makes it difficult to find one of the simplest ways to perform any given activity.
On this in-depth tutorial, which I offered at PyCon 2019, you may use pandas to reply questions on a real-world dataset. By means of every train, you may study necessary information science expertise in addition to “finest practices” for utilizing pandas. By the tip of the tutorial, you may be extra fluent at utilizing pandas to appropriately and effectively reply your individual information science questions.
“This tutorial is among the best programs on pandas, if not the very best, particularly for individuals who do not have superior degree in pandas. I’ve been utilizing pandas for a while however I found issues on this course that have been superb for me.” – S.R.
That is an intermediate degree tutorial, so if you happen to’re new to pandas, I like to recommend beginning with my different video collection: Easier data analysis with pandas.
If you wish to observe together with the workout routines at house, you may download the dataset and notebook from GitHub.
Listed below are a number of the matters lined within the video:
- adjusting for bias in your dataset
- dealing with lacking values
- selecting an acceptable plot
- customizing your plot
- utilizing the datetime information sort
- filtering utilizing loc versus question
- utilizing a number of aggregation features
- checking for small pattern sizes
- methodology chaining
- verifying your outcomes utilizing random samples
- evaluating a “stringifed” Python container
- making use of a customized operate to a Collection
- writing lambda features
Let me know in case you have any questions, and I am glad to reply them!
P.S. In the event you like this video, it’s best to take a look at my interactive pandas course, Analyzing Police Activity with pandas.