Forum Discussion
How to represent a variable's variance over time
Hi!
Let me see if I can explain what my question is:
Once a week, I receive a database, with aggregate estimates of future demand for air transportation (by airline, country, airport, etc) from that current moment in time until 9-12 months into the future. I therefore have several versions of what demand is forecasted to be for the rest of the whole year, during the year, depending on which week of that year I am looking ahead from.
I want to be able to analyze if that demand is on an increasing or decreasing trend during the next months, based on the differences in volume reported for those months in previous weeks.
I'm having trouble showing this in an easy way through a chart, since I have several versions for the same month and several months (not to mention other variables I can filter, such as airline or airport). I would need something like a box and whiskers chart based not on statistical relationships but time, in which the several whiskers would not represent quartiles but successive moments in time, progressively more recent.
Anyone have any ideas on how to best represent this?
I hope I explained myself well.
My current database looks something like this:
Report Year | Report Week | Country | Origin Airport | Destination Airport | Airline | Operating Year | Operating Month | Flight Frequency | Volume
Use a scatter plot with play axis.