Introduction to Data Science in Python
Preface
Python workshop taught by Professor Markus Loecher (HWR Berlin). The workshop will run in-person June 6 - 9 in room VC 14-280 (from 10am to 1pm and 2 to 5pm).
About this python workshop
Goals
This 4 day workshop is intended to introduce participants to the python language. It is designed to provide the solid foundation needed to conduct data analysis and visualization for data science. While no previous experience is required, some basic programming or data science experience is helpful.
I will lean heavily on the book Python for Data Analysis (as well as the Python Data Science Handbook).
The first day will focus on the fundamentals of data types and flow structures while the ultimate goal of the course will be to introduce you to statistical thinking, data literacy and modeling.
DataCamp
DataCamp is a pretty good resource for students to learn coding and data analysis skills. By completing the DataCamp courses listed below we would be able to significantly shorten the time we spend on basics and open up more space for data science concepts.
If you have extra time: * Data Manipulation with pandas
And much more advanced and totally optional:
Coding Environment
The most convenient environment for you to code in might be Google Colab, for which you probably need a gmail account. It does not hurt to look at the 2-minute intro video. If you prefer a real IDE, I would recommend Visual Studio or PyCharm. (I will not be able to help much with the latter though)
Agenda
Day 1: Basic python programming
- basic data types: lists, tuples, dictionaries, strings
- control structures (for, if else, while)
- functions
- numpy arrays: slicing and subsetting, axis
- Probabilistic Simulations
- basic plots
Day 2: pandas and visualization
- pandas Data Frames: slicing and subsetting
- Counting and Summary Statistics
- Handling Files
- Grouped Operations
- plotting with pandas
- Contingency Tables as models
Day 3: Statistical Modeling
- A/B Testing and sampling distributions
- Hypothesis Testing
- parametric
- permutation
- the bootstrap
- regression
- simple and multiple
- logistic
- categorical variables and interactions
- regularization
Day 4: Machine Learning
- Basic ML tools
- Cross Validation
- sklearn
- Data Cleaning
- Classification and Regression Trees
- Random Forests and Boosting
- Exlainable ML
- Partial dependence plots
- SHAP values
If time permits
- Word embeddings such as word2vec
- Sentiment analysis
- Internet scraping
- Topic models
Lecturer
Professor for Mathematics and Statistics