This project explores the 2018 Central Park Squirrel Census dataset.
The goal was to clean the data, handle missing values, and analyze squirrel sightings to uncover patterns in their activity.
-
Data Cleaning
- Dropped nearly empty columns (
Hectare Conditions Notes,Litter Notes). - Filled missing categorical values (
Litter Notes,Other Animal Sightings,Hectare Conditions). - Filled missing numerical values (
Total Time of Sighting) with the mode. - Converted
Datefrommmddyyformat into a proper datetime and extracted the Day.
- Dropped nearly empty columns (
-
Exploratory Data Analysis (EDA)
- Visualized top 10 hectares with most squirrel activity.
- Compared AM vs PM sessions (50/50 split).
- Checked hectare conditions (
Busy,Calm, etc.) and their squirrel counts. - Analyzed daily trends → peak activity on Oct 13, 2018.
- Looked at litter levels and squirrel presence.
- Compared sightings by number of observers (sighters) using bar and scatter plots.
- Most squirrel sightings were in Busy and Calm areas.
- Sessions were evenly balanced between AM and PM.
- October 13th had the highest squirrel activity (430+ sightings).
- Most data came from single observers, though group sightings showed higher variation.
- Litter presence (
None,Some,Abundant) was recorded but didn’t dominate the dataset.
- How to handle missing values using drop, fillna, and mode.
- Converting and working with dates in pandas (
to_datetime,dt.day). - Creating meaningful visualizations (bar plots, pie charts, line plots, scatter plots).
- The importance of documenting analysis with markdown + comments so results tell a story.
- How structured notebooks can feel like mini research reports when organized properly.
- Explore relationships between squirrel behavior and weather data.
- Model daily activity trends to see if there are predictable patterns.
- Use clustering to group squirrels by behavior and location.
If you notice any mistakes or have suggestions for improvements, feel free to open an issue or drop feedback — I’d love to learn and make this better!
👩💻 Author: Bushra 🪐