Homework Types

Table of Contents

Code

A coding homework represents a complete and functional application solving a particular problem or implementing a business logic. The students should follow the coding best practices of the language being used.

Style guide by language:

• Python style guide
• Scala style guide - EPFL
• Scala style guide - Databricks

Formatting tools:

• Black for Python
• Scalariform code formatter

A complete code-homework should contain both in-code documentation, documentation files, and tests.

• Total points: 25

Essay

The number of words in the essay should satisfy the following constraint:

$$700 \leq words \leq 1000$$

Students should follow the “five-paragraph essay” format, which consists of three principal components:

• Introduction - The principal purpose of the introduction is to expose the thesis statement explicitly. An effective introductory paragraph usually begins with a hook to grab the reader’s attention. This might be a personal anecdote, relevant statistics, or a quote. After that, the writer should move on to a clear, one-sentence thesis statement.
• Body - A good essay usually contains roughly three body paragraphs. The first sentence of these paragraphs is known as the topic sentence and should provide an argument in favor of the thesis. Body paragraphs make use of examples, supporting evidence, statistics, and citations. To ensure continuity, you might want to use a “transition phrase” to connect different paragraphs (e.g., “moreover”, “on the other hand”, “by contrast”, “furthermore”).
• Conclusion - This paragraph should start with a “concluding transition” and briefly recap on the main arguments of the essay. Lastly, it should restate the thesis statement.

You can read more recommendations here.

Consider the following aesthetic hints:

• Are your sentences elegant and careful?
• Have you chosen each word properly?
• Does the sentence say what it is supposed to say?
• Do your paragraphs constitute the elaboration of a single idea?
• Are they sufficiently comprehensive and concise?
• Does the essay succeed as a unit?
• Does it make an identifiable and intelligent statement?

The grading of the essay is simple:

Total points: 15

Grammar examination includes, but is not limited to:

• Spelling, punctuation, and syntax.
• Correct use of sentences (e.g., no sentence fragments or run-on sentences).
• Consistent verb tense usage.

If you are interested in improving your writing skills, consider reading:

• The Grammarly blog. Contains relevant posts such as the top 10 grammatical mistakes and 15 easy steps to improve writing.
• The 10-step process for stronger writing from BigThink.

Report

The data-analysis reports must be done using Python in rmarkdown or jupyter notebook / lab. The final deliverable should be a PDF file with the related source (.Rmd or .ipynb).

The minimum structure of the report should include:

1. Executive summary - Explain the problem, the dataset, and your findings.
2. Body - The body contains the development and technical aspects of the report. It’s usually “question oriented”. The following sections are expected:
   • Data - Description of the data source, format, and shape. For each relevant variable, the type should be identified. How does the data quality looks? Does it includes some missing values?
   • Univariable analysis - Analyze the distribution of the target variable. Does the target variable follows a normal distribution? Should we apply a custom transformation?
   • Multivariable analysis - Identify the relevant variables in the dataset. This can be done by a correlation analysis. Visualize results (e.g., correlation matrix, scatterplot, box-plots). What’s the distribution of the related variables? What variables do you think can be used as predictors?
3. Conclusions - Closing paragraph that contains the most relevant insights of the exploratory analysis.
4. Appendix - Any other resource not directly related to the main analysis but of further interest. This might include:
   • A more in-depth description of a technical procedure.
   • Detailed output of a process.
   • Mathematical demonstrations.
   • Supporting code.

Total points: 10

If you are interested in improving your data-analysis skills, consider reading some kaggle kernels. Particularly this comprehensive data exploration with Python.

Exercise

Total points: 10

The final score is proportional to the number of correct answers. Should be delivered via github.

Data Entry

Total points: 5

Data entries will be evaluated as binary.