New York Vehicle Collisions (2014—2023)
Detailed data on traffic collisions in New York between 2014 and 2023, useful for modeling, prevention, and statistical analysis.
Description
This dataset covers all traffic accidents reported by police in New York between 2014 and 2023. It includes over 2.9 million entries with detailed information on the date, time, location, victims, vehicles involved, and contributing factors. This data comes from official forms such as MV104-AN and is updated regularly.
What is this dataset for?
- Analyze accident trends in the various geographic areas of NYC
- Training predictive models for collision prevention
- Evaluate human, technical, or environmental factors related to accidents
Can it be enriched or improved?
Yes, data can be cross-referenced with external sources such as weather, traffic or road infrastructure. You can also correct missing values, add categorizations (holidays, rush hour, etc.) or create temporal aggregations.
🔎 In summary
🧠 Recommended for
- Urban analysts
- Road safety data scientists
- Open data projects
🔧 Compatible tools
- Pandas
- Spark
- Table
- BigQuery
- Scikit-learn
💡 Tip
Filter out serious or fatal accidents for targeted analyses, and use GPS coordinates to map black spots.
Frequently Asked Questions
Can this data be used in a commercial context?
Yes, the CC0 license allows unrestricted commercial use.
Does this dataset contain personal data?
No, all data is aggregated by collision event and does not contain any personal information.
How do I deal with missing values in the file?
It is recommended to delete rows that are too incomplete or to impute missing values depending on the analysis context.
