Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 6 years ago.
Improve this question
I am starting a new project and I have the problem with the measurability of the requirement usability.
With is an good atomic measurable indicator for usability?
Number of klicks or keystrokes per action, weighted with the estimated relative frequency of the action.
However, this metric does not take into account a few important aspects of usability:
How easy it is to memorize the sequence of clicks and keystrokes
The amount of feedback provided to the user
The likelyhood of invalid inputs
Still it can be useful for comparing user interfaces that are similar with respect to these aspects.
Related
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 6 days ago.
Improve this question
I apologize if the title is a little confusing, but hopefully that highlights the point of my question. If a method has recursively called itself multiple times and some base case is reached that causes all recursive calls to return/'undo', what would be a proper term to describe that?
Exiting, undoing, and unraveling recursion all come to mind, but I'm mostly wondering if there's an agreed upon term for it that I can use in comments. Thanks!
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 2 years ago.
Improve this question
When should one consider using a metamodel in place of an expensive analysis component? I haven't actually seen an example of an openMDAO project that uses metamodels yet, so I was wondering if they're recommended or whether it's still best to treat expensive analyses like normal ones.
Using a metamodel is a trade-off. You trade accuracy vs execution time so it is up to you in your context.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 4 years ago.
Improve this question
What were the design decisions that led to R having often more than one way of doing things, that have subtle difference? See, for a good example,
https://www.r-bloggers.com/r-na-vs-null/
More more such issues are here, some which are justified, some which are not http://r4stats.com/articles/why-r-is-hard-to-learn/
From a software engineering perspective, having such choices in a language screams for having subtle and hard-to-find bugs in your code (e.g. in Python the whole point of writing "pythonic" code, that avoids ambiguity and is easy to read and consistent in style). So there must be some major advantages of having that. What are they?
Closed. This question is off-topic. It is not currently accepting answers.
Want to improve this question? Update the question so it's on-topic for Stack Overflow.
Closed 11 years ago.
Improve this question
I'm looking to implement an ELO ranking system. I've read the wikipedia articles and I'm confused about the start rank for players who enter the system at a later point. The common solution is to use a provisional ranking system but I'm curious if anyone can point me to specific numeric details:
what K value do new players get?
how long does a player stay in provisional mode?
how does K value change as rank changes?
I'm sure there are many variations, I'd just like to know actual numbers for a system that someone has implemented successfully.
Thanks for your time.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 6 years ago.
Improve this question
I heard this a lot when talking about software engineering and abstract data types, what does this do? Can anyone give me a concrete example of this concept?
A representation invariant is a condition concerning the state of an object. The condition can always be assumed to be true for a given object, and operations are required not to violate it.
In a Deck class, a representation invariant might be that there are always 52 Cards in the deck. A shuffle() operation is thus guaranteed not to drop any cards on the floor. Which in turn means that someone calling shuffle(), or indeed any other operation, does not need to check the number of cards before and after: they are guaranteed that it will always be 52.