1
$\begingroup$

I've collected bus arrival times at my local bus stop from the past month - so I have every time my bus (a specific bus number) shows up at my bus stop for each day of the week (Monday, Tuesday, etc.).

I am struggling at determining the best and clearest way to display this data. Eventually I want to use a clustering algorithm to help understand the most likely time the bus will show up. So on a Monday I know if the bus is more likely to show up at 7.45am or 7.48 am.

I believe 8 charts will be best - one chart for each day of the week, and then one final chart that shows the average regardless of day of week. What would be my best chart type for clearly visualising this data?

$\endgroup$
0
$\begingroup$

How about using a boxplot, so you can also visualize the variance to some degree. Each boxplot could represent one weekday.

When you look only at the mean (and be at the busstop at this time) you have a good chance to miss the bus (about 50% for a well-shaped non-skewed distribution).

You want to make sure that you be at the stop at the right time (95% chance of getting the bus or so). Therefore variance is important.

enter image description here

The image is from this page: https://www.datamentor.io/r-programming/box-plot/

Alternatively, you could train a model on the arrival time, make predictions for each day, and plot the predicted value including a 95% confidence interval. But this probably is too much here.

$\endgroup$
  • $\begingroup$ The box plot could work very well for once i determine averages for specific times for each day - for example, the data is as follows: 7.43, 7.40, 7.45, 12.00, 12. 04, 11.59, 13.00, 13.01, 13.01 for three days. So I know there should be three clusters - but I am not sure is the best way to determine/implement finding the average at ~7.44ish $\endgroup$ – C Murphy Nov 16 at 16:26
0
$\begingroup$

Stacked histograms, or stacked Kernel Density Estimates.

Use a different color per day. Then you'll easily be able to compare days, e.g., see that the Sunday is missing at certain times while the stripes for monday to friday are the same.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.