Support other increments of size

[mwermelinger]

Currently the size is always doubled, but for exponential or factorial algorithms it would be better to increment size by 1. For most flexibility, the functions would allow to increment by a fixed amount and/or by a factor. Arguments could be:

• start value/size
• steps number of doublings or additions, must be ≥ 0
• factor: a float ≥ 1, default 2
• increment: an int ≥ 0, default 0

The function then measures run-times for start, start*factor+increment, etc. for as many steps as given. (steps = 0 means it does only for start)

• update library (should check that factor > 1 or increment > 0, as otherwise size doesn't grow)
• update code guide
• update/add an example in an essay (maybe a student essay)
• update template

[densnow]

I've made a start on this and pushed a 39-support-other-increments branch.

So far I have:

• modified the library including code, docstrings and assertions
• modified the example and template essays by changing the double parameter to steps

I won't tick any of the boxes until I have had another look tomorrow and will continue working on it.