Hanze

Go back to the main page

Example Portfolio

The Portfolio Assignment

This document outlines the requirements and expectations for your portfolio, designed to showcase your data analysis skills. Think of this as a structured project where you’ll apply what you’ve learned to a real-world dataset. The provided example portfolio aims for a 7.5 level, but you can certainly aim higher!

Your Workflow: Step-by-Step

Follow these steps to complete your portfolio effectively:

Dataset Selection:

Search for datasets, for example Kaggle, and explore the vast collection of datasets. Choose one that genuinely interests you and allows for meaningful analysis relevant to topics covered. In particular, consider datasets from the field of Life Sciences but data sets from other fields are also allowed.

Crucially, once you’ve selected your dataset, provide its direct link in the Excel file your instructor will provide. This step is vital to prevent duplicate datasets from being used by different students.

In order to fulfill all the requirements, you may need to use multiple datasets (see below).

Structure of the Portfolio:

Adhere strictly to the example portfolio’s (see link above) structure. This ensures consistency and helps you present your analysis logically and comprehensively. Don’t deviate from this structure unless explicitly instructed or if you have a compelling, justified reason.

Keep the following in mind:

Directory Hierarchy:

Let’s take a look at the Example Excel portfolio. The directory hierarchy is as follows:

example_portfolio_excel/
├── analysis_1_calories_food/
│   ├── checklist_data_analysis_portfolio.xlsx    ← workflow checklist
│   ├── data/
│   │   ├── nutrients_csvfile.csv                 ← raw dataset
│   │   ├── nutrients_csvfile_modified.csv        ← cleaned dataset
│   │   └── source.txt                            ← notes / provenance
│   └── analysis/
│       ├── calories_food.xlsx                    ← main Excel analysis (formulas, charts)
│       └── calories_food.docx                    ← report/documentation of findings
│
├── analysis_2_zomato_stock_price/
│   ├── checklist_data_analysis_portfolio.xlsx
│   ├── data/
│   │   └── zomato.csv                            ← stock data
│   └── analysis/
│       ├── zomato_stock_price_analysis.xlsx      ← Excel analysis
│       └── zomato_stock_price_analysis.docx      ← report

Likewise, the zip file for R has the following directory structure:

example_porfolio_R/
├── analysis_1_calories_food/
│   ├── checklist_data_analysis_portfolio.xlsx   ← project checklist
│   ├── data/
│   │   ├── nutrients_csvfile.csv                ← raw dataset
│   │   ├── nutrients_csvfile_modified.csv       ← cleaned dataset
│   │   └── source.txt                           ← data source notes
│   └── analysis/
│       ├── Calories_food.Rmd                    ← R Markdown analysis script
│       ├── Calories_food.html                   ← rendered report
│       ├── .Rhistory                            ← R command history
│       └── pics/
│           ├── fig01.png
│           └── fig02.png                        ← plots/visuals used in report
│
├── analysis_2_zomato_stock_price/
│   ├── checklist_data_analysis_portfolio.xlsx
│   ├── data/
│   │   └── zomato.csv                           ← stock price dataset
│   └── analysis/
│       ├── zomato_stock_price_analysis.Rmd      ← R Markdown analysis script
│       ├── zomato_stock_price_analysis.html     ← rendered report
│       ├── .Rhistory
│       └── pics/
│           ├── fig01.png
│           └── fig02.png

Tip: Have a look at the example portfolio zip files to understand the directory hierarchy.

You can use as many datasets as required to meet all the requirements for the checklist (see below). Number your analysis like the example above. If you need to modify your dataset (for example to deliberately duplicate a row to demonstrate data cleaning), make sure that you rename it filename_modified.csv. This clearly indicates that you made modifications to the source file. These modifications should be explained and justified in the corresponding analysis text.

The Checklist

You can download the checklist (see link above) from this page. Go through each item on this list meticulously. Note a 1 (one) if the corresponding item is present and a 0 (zero) if not. For every row, not only check it off but also clearly justify within your portfolio how you’ve addressed that requirement. At the end of each analysis, write a dedicated section to justify the checked items and show a screenshot. This demonstration of compliance is essential for assessing your portfolio by your instructor.

An example:

The first analysis is about the calories in various food products and can be found within the subfolder analysis_1_calories_food. The last paragraph is about the checklist. Put a screenshot of the checklist and also clearly describe what items are missing:

cheklist

As you can see, the clustered bar plot and the XY-scatterplot are not included in the first analysis. The main reason for this is that the dataset was not suitable for this type of analysis. Therefore a second analysis was performed. The analysis can be found within the subfolder analysis_2_zomato_stock_price. Note that the name of the subfolder starts with analysis_2. This way, the folder will show as a second item in a file explorer. The last paragraph concerns the checklist again.

cheklist

Keep in mind that data loading and cleaning needs to be described for each analysis as datasets may vary in quality.

Note that the checklist is check marked cumulative. This means that the previous checklist is used as a starting point to check the last two items (Clustered barplot and XY-scatterplot). This works well only if the analysis folders are also named in a logical manner:

If you organize your work in a similar logical way, your instructor will be able to easily determine if all the requirements are met before grading your portfolio.

If you do not adhere to these guidelines, your portfolio will not be graded!

Relative filepaths

Use relative filepaths in your project folders for the analysis in R. The primary reason to use relative filepaths is to ensure portability of your project. Also, use OS-agnostics filepaths so that your analysis can be run on Windows, Mac and Linux computers (all these different OSes are used by your instructors and definitely later in your research team).

Example:

Wrong (absolute path Unix system like Mac OS or Linux):

/mnt/data/example_portfolio_R_extracted/example_porfolio_R/analysis_1_calories_food/data/nutrients_csvfile.csv

Wrong (absolute path Windows system):

C:\Users\<YourUsername>\Downloads\example_portfolio_R_extracted\example_porfolio_R\analysis_1_calories_food\data\nutrients_csvfile.csv

Right (relative path for Windows and Unix system like Mac OS or Linux):

../data/nutrients_csvfile.csv

Also make your path OS agnostic.

Wrong (will break on Unix system like Mac OS or Linux):

..\data\nutrients_csvfile.csv

Right (will work on all listed systems):

../data/nutrients_csvfile.csv

Even better: use the file.path() function. Example code:

data_path <- file.path("..", "data", "nutrients_csvfile.csv")
data <- read_csv(data_path)

Self plagiarism

As both the Word file and the Rmd file contains the analysis of the same dataset, self plagiarism is allowed.

Example Word:

cheklist

Example in Rmd:

cheklist

Perform the same analysis in Excel and R. Small deviations are allowed (like the use of violin plots or other plot types in R that are missing in Excel or visa versa) but the general ouline of your analysis should be similar. In this way, you can also compare the outcome between Excel graphs and graphs ceated in R. The example portfolio can also clarify this.

Use of LLM (large language models/generative AI)

Some use of generative AI is permitted but do not rely to heavily on it. It is allowed for debugging little mistakes in your code and correcting and rephrasing small paragraphs. What is not allowed is to use LLM to do entire analysis and create Rmd files from prompts. The resulting code often has a typical “signature” and if instructors suspect that entire blocks of codes or large parts of the text is AI generated, an interview (involving two instructors) may follow (see below).

Stronger Narratives and Interpretations

Don’t just show plots; tell a story with your data. Clearly articulate what each visualization reveals, connect findings across different analyses, and draw insightful conclusions that directly address potential research questions or hypotheses. Robust Justification: Consistently explain why you made certain decisions (e.g., why you chose a particular method for handling missing data, why a specific plot type was best suited for a certain variable). By putting in the extra effort and thoughtful consideration, you can significantly enhance your portfolio and demonstrate a deeper understanding of data analysis principles.

What if your dataset does not support specific skills?

Imagine that your dataset is already clean and does not require any cleaning. In that case show what you would have done if missing data was present. Also show how you would handle duplicate rows etc. You can modify your raw data (see above, make sure you rename your file) and show how you would handle missing data and duplicate rows and missing column headers et cetera.

Beyond the Basics

The example portfolio is a solid foundation, aiming for a 7.5. To achieve an even higher score, consider these enhancements:

Make it beautiful

You can make your R Markdown report in HTML more beautiful by changing the standard YAML front matter at the top of an R Markdown (.Rmd) file. It declares metadata and tells R Markdown how to render the document. Here’s a quick breakdown and a few small tips. Example YAML front matter:

---
title: "Calories Food"
author: "Jane Doe"
date: "2025-04-10"
output: 
  prettydoc::html_pretty:
    theme: cayman
    highlight: github
---

Note that the YAML front matter code block should start and end with 3 dashes.

What each line means

What this does in practice

When you knit the .Rmd (e.g., click Knit in RStudio R Markdown will create an HTML file styled by prettydoc with the Cayman theme and GitHub-like code highlighting. If prettydoc isn’t installed, the render will fail until you install it.

install.packages("prettydoc")

If you do not like the Cayman theme you can choose some others like:

Theme name Visual style summary
cayman Clean, minimal GitHub-Pages look with blue header. (default)
architect Crisp white layout, bold black title bar; inspired by “Architect” Jekyll theme.
leonids White body, black banner; small-caps title; subtle typography.
tactile Beige background, warm tone, retro “notebook” feel.
hpstr Modern, dark header, sans-serif fonts, slightly higher contrast.

Submission

Once you’re confident that all components are complete and polished, zip your folders (as two separate zip files) and upload and share your zipped portfolio via BrightSpace using the corresponding Excel and R links.

cheklist

Review and Feedback:

Your instructor will first evaluate if all the requirements are met based on your portfolio and the checklist. If this is the case, your instructor will evaluate your portfolio using the rubric shown below. Make sure you’re familiar with this form beforehand to understand the grading criteria. Finally, you’ll receive your graded rubric via Brightspace. If your instructor suspects the work submitted relies too heavily on AI assistance (beyond code corrections, paragraph edits and rephrasing), it could lead to an investigation via CGI (Communication, Discussion, insight interview). Two instructors will be involved in the CGI.

Interview regarding the suspected misuse of an LLM

As described above, if your instructor suspects that your work relies too heavily on the use of generative AI, you will have a brief 15-minute interview session with your instructor and a second instructor. If it turns out that you are not able to explain the process of your analysis and code, the portfolio will not be evaluated any further.

Rubric (grading scheme)

The Rubric is organized as follows:

0 points 2 points 3 points 4 points 5 points
Not done Unsatisfied Satisfied for basic analysis Good analysis and proper conclusions and description analysis beyond the basics and excellent descriptions

The items that are scored (both and Excel and R) are:

The weight of these items is not equal. The weights of the corresponding items and the scaled points can be found in the table below:

Item Weight Scaled Points (max)
Data Import 2 10
Data Cleaning 1 5
Data Analysis 3 15
Data Visualization 4 20

You can find the complete rubric on Brightspace.

The grade for Excel and R will be calculated as follows:

\[\text{Grade} = \max\left(1,\frac{\text{aquired points}}{\text{total points}} \cdot 10 \right)\]

You will receive a grade for Excel and a grade for R. The Excel part counts for 40% and the R for 60%.
So your end grade will be:

\[\text{End grade} = 0.4 \cdot \text{Excel grade} + 0.6 \cdot \text{R grade}\]

Please keep in mind that Osiris calculates the final grade. The instructors enter separate grades for Excel and R. Since these are two submodules, both components must be a passing grade. This is a general rule for how Osiris calculates final grades at Hanze.

Good luck! 🍀

Go back to the main page
⬆️ Back to Top

This web page is distributed under the terms of the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Creative Commons License: CC BY-SA 4.0.