3 ms·

> maps (with arbitrary scalar keys, not just strings) Floating point keys just seem like a terrible idea.

dragonwriter 11y ago
> Floating point keys just seem like a terrible idea. I don't see why floating point (with a defined precision) keys are any worse fundamentally than any finite-domain scalar keys. The applications for which they are useful may be limited, and they may be the wrong choice for some uses, but that's true of anything.
- jimktrains2 11y ago
  > floating point (with a defined precision) If it's fixed-precision it's not floating point, it's basically an integer we interpret with a decimal point somewhere when we display it. If you didn't mean fixed-precision, then I'm not sure how actual floating points could sanely act as a hash key.
  - tel 11y ago
    Even with a defined precision they're weird in that they "ought to" be modeling the reals and therefore lack computable equality. Which makes it hard to write a well-defined hash.
  - dragonwriter 11y ago
    > If it's fixed-precision it's not floating point I'm not referring to fixed point, but fixed-size representation. > If you didn't mean fixed-precision, then I'm not sure how actual floating points could sanely act as a hash key. I don't see what the problem is. Certainly, there's problems if you do certain floating point calculations and blindly look for the result as a key in a mapping, but that's not a problem with floats as keys, its a problem with that particular usage pattern.
    - jimktrains2 11y ago
      > I'm not referring to fixed point, but fixed-size representation. I'm still not sure what you mean as any IEEE floating point (which in reality all anyone uses) is fixed-size. If you decide to truncate, round, do w/e to this value, you may still not get exactly the same answer for similar computations or even the same computation on different machines. There isn't a reason you can't (well, shouldn't) do float_val == 0.0 -- it won't work in all circumstances. Please explain your method of making sure float_val == 0.0 while retaining the essence of being floating point (i.e. not using fixed-point).

> Floating point keys just seem like a terrible idea. I don't see why floating point (with a defined precision) keys are any worse fundamentally than any finite-domain scalar keys. The applications for which they are useful may be limited, and they may be the wrong choice for some uses, but that's true of anything.

jimktrains2 11y ago
> floating point (with a defined precision) If it's fixed-precision it's not floating point, it's basically an integer we interpret with a decimal point somewhere when we display it. If you didn't mean fixed-precision, then I'm not sure how actual floating points could sanely act as a hash key.
- tel 11y ago
  Even with a defined precision they're weird in that they "ought to" be modeling the reals and therefore lack computable equality. Which makes it hard to write a well-defined hash.
- dragonwriter 11y ago
  > If it's fixed-precision it's not floating point I'm not referring to fixed point, but fixed-size representation. > If you didn't mean fixed-precision, then I'm not sure how actual floating points could sanely act as a hash key. I don't see what the problem is. Certainly, there's problems if you do certain floating point calculations and blindly look for the result as a key in a mapping, but that's not a problem with floats as keys, its a problem with that particular usage pattern.
  - jimktrains2 11y ago
    > I'm not referring to fixed point, but fixed-size representation. I'm still not sure what you mean as any IEEE floating point (which in reality all anyone uses) is fixed-size. If you decide to truncate, round, do w/e to this value, you may still not get exactly the same answer for similar computations or even the same computation on different machines. There isn't a reason you can't (well, shouldn't) do float_val == 0.0 -- it won't work in all circumstances. Please explain your method of making sure float_val == 0.0 while retaining the essence of being floating point (i.e. not using fixed-point).