2236
DF General Discussion / Re: DF has a SHALLOW learning curve
« on: May 19, 2014, 11:02:57 pm »
The real nail in the coffin though is that you can even have graphs that turn back on themselves or have areas swept out and not just lines, or all manner of other shenanigans, in special circumstances.
For example, actual effective taxi fares amongst people who are in too much of a hurry to wait for change will have weird, non-line area shapes on such a graph, because they will depend on randomly varying factors like "what types of bills customers happened to have in their pockets at the time." And these would even be lawfully predictable, based on bill carrying statistics, despite making absolutely no sense as an ordered function.
Yet makes absolutely perfect sense when you treat it like what it is: a correlation plot of pairs of observed or predictable values. Not a function.
For example, actual effective taxi fares amongst people who are in too much of a hurry to wait for change will have weird, non-line area shapes on such a graph, because they will depend on randomly varying factors like "what types of bills customers happened to have in their pockets at the time." And these would even be lawfully predictable, based on bill carrying statistics, despite making absolutely no sense as an ordered function.
Yet makes absolutely perfect sense when you treat it like what it is: a correlation plot of pairs of observed or predictable values. Not a function.