Talk:Quickselect algorithm: Difference between revisions

Content added Content deleted
(→‎How is pivot chosen?: added my experience with RANDOM versus a midpoint (for a selection of a pivot). -- ~~~~)
m (→‎Task Testing Values: added a possible addition. -- ~~~~)
Line 20: Line 20:
In future tasks I suggest you to propose common task data values like [90, 80, 70, 60, 50, 0, 10, 20, 30, 40] instead of [9, 8, 7, 6, 5, 0, 1, 2, 3, 4] because with the first it's a little less easy to produce a correct output in presence of code bugs. Also to reduce the probability of having buggy entries that give a correct answer on just the common task data values, I suggest to offer more test cases.
In future tasks I suggest you to propose common task data values like [90, 80, 70, 60, 50, 0, 10, 20, 30, 40] instead of [9, 8, 7, 6, 5, 0, 1, 2, 3, 4] because with the first it's a little less easy to produce a correct output in presence of code bugs. Also to reduce the probability of having buggy entries that give a correct answer on just the common task data values, I suggest to offer more test cases.
:It is a compromise. I had to give some values from when the task was created and chose the ones I did to at least have a common set of values that were easy to check by eye. At the time of writing, I wanted something, rather than nothing at all. --[[User:Paddy3118|Paddy3118]] ([[User talk:Paddy3118|talk]]) 13:14, 11 October 2013 (UTC)
:It is a compromise. I had to give some values from when the task was created and chose the ones I did to at least have a common set of values that were easy to check by eye. At the time of writing, I wanted something, rather than nothing at all. --[[User:Paddy3118|Paddy3118]] ([[User talk:Paddy3118|talk]]) 13:14, 11 October 2013 (UTC)

: I also would suggest adding another part to the task requirement:   ''Ensure (but do not show) that the program works with a single value.''     -- [[User:Gerard Schildberger|Gerard Schildberger]] ([[User talk:Gerard Schildberger|talk]]) 10:56, 19 November 2013 (UTC)