6 Ending Off with a Few Thoughts

We started with the idea to extract a number of informative aggregate characteristics of roads in India by state/UT, but found that there may be some issues with accuracy (coordinate system and trimming near state/UT boundaries) and computation time (time it took to process one state/UT).

Ideally, we should project each state/UT to an appropriate planar coordinate system to achieve more accurate spatial intersection results. There should probably also be a good amount of trimming/cutting involved as roads tend to be partially outside the boundaries of the states/UTs.

Although the data was only a few megabytes, the computation time was roughly 10.94 secs minutes for the single state/UT, estimated to be 393.73 secs for all state/UTs. This could be problematic as we receive finer boundary or road data, and scale to larger more detailed datasets.

Some exploration into optimizing the spatial intersections and trimming/cutting could be a good start to obtaining quicker results in a reasonable amount of time (especially for learning or education).

Getting the road lengths and vertices statistics could be useful (as far as I can tell) in providing information on road complexity and development. This could be a potentially good measure for getting a glimpse at the general road design at a state/UT level, while providing some insight into whether or not road design at that level has an effect on important applications such as road traffic safety.

Overall, R has proven to be great for trying out ideas, experimenting with new libraries, and providing an easier transition into open source data analysis for those unfamiliar with spatial data handling - particularly when install.packages just works without any hiccups on windows!

I am hoping to take this idea a little further and apply it to study traffic crashes in India, and perhaps refine this walkthrough so it becomes a little more beginner friendly.

It was a lot of fun to write this and I’m hoping to try out some more ideas/thoughts when I have time!