A repository for the course material used in a Data plus project training at Duke, which focus on visualization in python. Materials are free to use under MIT Liscence.
-
Session 1 Programming Basics:
-
Basics of programming, example used: Generating Fractals Using Chaos Game
-
Conditions, Functions, Recursions, Lists, Dictionaries
-
Basics of Dataframe, And Dataframe statistics
-
Basic Plotting, Scatter Plot
-
-
Session 2: Data Acquisition
-
HTML CSS Javascript Basics; How browser renders web content? Things covered in the example scrapping
-
Beautiful soup: How it Cleans the HTML tags and Structurize the document
-
Basics about the CSS selector
-
How to srape html table
-
Using Selenium for scraping in more complex scenarios
-
-
Session 3: Data Exploration
- Histogram Barchat, Time data formatting used in visualization (One Staition 20 days, line oscilations, inject scattering missing data, create some illegal data)
-
- Load the file and count the nas of each variables (Exercise)
-
- Illegally values (Exercise)
- Grouping using time hours into days, days into months
-
Session 4: Geographical Visualization
- Geopandas, geoplot:
- Illustrating grouping using the incentive of aggregating information
- Animation? Plotly
-
Session 5: Machine Learning
- Big categories of Machine Learning: Supervised/Unsupervised/Semi-Supervised
- Maximum Likelihood and loss function
- Random Forest, Naive Bayes, Linear Discriminent Analysis, SVM
- Linear, Polynomial Regression, Non Parametric Regression, Gaussian Process Regression
