agchesneau/trains_project

Project

For this mid-bootcamp project, I am asking the following question : "how is the reliability of French national train lines affected by the cities of origin and destination ?"

Data

For this project I used 2 datasets from https://ressources.data.sncf.com/pages/accueil/ : the regularity of "intercités" (regular speed) and TGV (high-speed). I cleaned the data, checked the presence of a relationship between line and reliability/number of trains canceled using a chi² contingency table, exported it to a csv file and analyzed it in Tableau (https://public.tableau.com/views/mid-project/Dashboard1?:language=en-GB&publish=yes&:display_count=n&:origin=viz_share_link).

Conclusion

I have found that high-speed trains tend to be more canceled than intercités. Also they are not both present equally over the land, some regions have more one or the other. The line which are less reliable (because of late or canceled trains) are pretty long and a lot of them are international which could have an impact also.

This project was purely intended to practice cleaning and analysing a dataset and is not a serious analysis of French train lines as the dataset are themselves aggragated dataset and cannot be used to make conclusion about overall reliability of trains, as stated in the information section (https://ressources.data.sncf.com/explore/dataset/regularite-mensuelle-tgv-aqst/table/?sort=date).