Date:

December 2023

TEAM:

Solo Project

CONCEPTS AND TOOLS USED:

  • Python

  • Markdown

  • Jupyter Notebook

  • Adobe Premiere Pro

  • Beta PDF

  • Bootstrapping

ABOUT:

I modeled botanical data in Python to test hypotheses about probability and randomness in nature via a case study into the classic “Loves Me, Loves Me Not…” flower game. The game involves stripping a flower of its petals one by one, with the final petal telling the player whether their love is requited. While it has long been regarded as a true test of the randomness of the universe, recent findings suggest flower petals form natural occurrences of the Fibonacci sequence and thus tend to possess an odd number of petals, indicating that the game may be biased towards an affirmative result. So, is “Loves Me, Loves Me Not…” a game designed only for optimists? I set out to test this theory by collecting real-world data from my surroundings, modeling permutations of petal formations in Python, and producing an informational video and a Jupyter notebook to display my dataset and findings to those who share my curiosity about data and the mysteries of the universe.

RESEARCH:

I opted to test whether this affirmative bias truly exists, and ultimately question the power of fate, through a flower-plucking field study of multiple floral species and deduce whether the “Loves Me Not” test is a truly random test of fate. Further, I examined additional variations of the game with multiple species of flowers to identify which flowers can increase one’s luck and which will leave true love up to fate. I counted over 250 flowers (marguerite daisy, rose, plumbago auriculata) and 7,283 petals to create the necessary dataset for my analysis.

HOW I BUILT IT / ANALYSIS:

I implemented a two-step analysis of bootstrapping followed by Beta PDF calculations. 

First, I implemented a Python algorithm for bootstrapping, a method that resamples a single data set to create many simulated samples, to estimate the sampling distribution of the probability of a particular flower having an odd number of petals. Then, I implemented a Beta PDF analysis. By modeling petal count as a Beta variable for each sample, I then calculated the probability of an odd number of petals via a formula for the expected value of a beta distribution. Finally, I published a thoroughly documented notebook written in Python and Markdown to share the statistical analyses in my report and present my dataset and petal-count projections.

Additionally, I included a script for users to generate their own random results from my dataset and test whether their affections are requited with the press of a button.

Feel free to take a look at my video summary (see below), written report, and code for a more in-depth view of my case study!

OUTCOME:

FUTURE DEVELOPMENT:

  • Implement animated UI for the flower simulator