Skip to content

Update "exact expt" #53

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 5 commits into
base: master
Choose a base branch
from
Open

Update "exact expt" #53

Changes from 3 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
bcd1ad3
Update exact expt
jpellegrini Sep 9, 2023
ded1b09
Fix typo 2.0 -> 3.0
jpellegrini Sep 9, 2023
e17f7c4
Include SCM
jpellegrini Sep 9, 2023
b4c11fc
Fix LIPS result for (expt 81 1/4)
jpellegrini Apr 7, 2024
f7af938
Fix thinko (inexact -> exact)
jpellegrini Apr 7, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
78 changes: 74 additions & 4 deletions surveys/exact-expt.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,9 +1,79 @@
# Exact expt

This tests whether exponentiation of exact numbers produces an exact result. The test expression was `(expt 1/3 3)`, except on systems without inexact rationals, where `(expt 3 3)` was substituted.
## Exact rational base

Exact exponentiation supported: Racket, Gauche, MIT, Gambit, Chicken with the `numbers` egg, Bigloo, Scheme48/scsh, Guile, Kawa, SISC, Chez, Vicare, Larceny, Ypsilon, Mosh, IronScheme, STklos, KSi, TinyScheme, Scheme 9, RScheme, S7, XLisp, Rep, UBM, Sizzle, Oaklisp, Sagittarius, Picrin, Owl Lisp, Chibi, Loko
This tests whether exponentiation of exact numbers produces an exact result.
The test expression was `(expt 1/3 3)`, except on systems without inexact rationals, where `(expt 3 3)` was substituted.

Only inexact exponentiation supported: plain Chicken, SCM, NexJ, JScheme, Unlikely, Elk, SXM, Dfsch, Inlab

No support or broken support for `expt`: SigScheme, Shoe, Mini-Scheme, SIOD, Schemik, Llava, FemtoLisp, LMU
| System | `(expt 1/3 3)` | `(expt 3 3)` |
|-------------|-----------------------|--------------|
| Bigloo | | 27 |
| Biwa | | 27 |
| Chez | 1/27 | |
| Chibi | 1/27 | |
| Chicken | 1/27 | |
| Elk | | 27 |
| Gambit | 1/27 | |
| Gauche | 1/27 | |
| Guile | 1/27 | |
| Kawa | 1/27 | |
| LIPS | 1/27 | |
| Loko | 1/27 | |
| MIT | 1/27 | |
| Racket | 1/27 | |
| Sagittarius | 1/27 | |
| Scheme48 | 1/27 | |
| Scheme9 | | 27 |
| SCM | 3.7037037037037035e-2 | |
| STklos | 1/27 | |
| Tinyscheme | | 27 |


## Exact rational exponent

What happens when the exponent is an exact rational, the base is exact, and it *would* be possible to return an exact result?

The following expression was evaluated:

```
(expt 81 1/4)
```

Implementations that do not support rationals are not included.
Some Common Lisp systems were also checked, for comparison.

| system | `(expt 81 1/4)` |
|-------------|--------------------|
| Chez | 3.0 |
| Chibi | 3.0 |
| Chicken | 3 |
| Gambit | 3 |
| Gauche | 3 |
| Guile | 3.0 |
| Kawa | 3.0 |
| LIPS | error |
| Loko | 3.000000238418579 |
| MIT | 3.0 |
| Racket | 3.0 |
| Sagittarius | 3.0 |
| Scheme48 | 3.0000000000000004 |
| SCM | 3.0 |
| STklos | 3.0000000000000004 |
|-------------|--------------------|
| ABCL | 3.0 |
| CCL | 3.0 |
| Clisp | 3 |
| ECL | 3.0 |
| GCL | 3.0000000000000006 |
| SBCL | 3.0 |

* The systems with result `3.0` in the first column are those where `(= 3.0 (expt 81 1/4))` is true.
In the other systems, which do not return the precise `3.0` floating-point representation, the
returned value indicates at least one digit where it differs from 3.0.

* The test `(= 3.0 (expt 81 1/4))` was necessary, since we do not know if `3.0` actually means a
precise representation of the number. For example, STklos will answer `3.0`, but after changing
the default print precision with `(real-precision 50)`, it will show that the number is not
the same as `3.0`.