Timeline for Different ways (and the fastest) to compute sines (and cosines) in Arduino
Current License: CC BY-SA 3.0
24 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 6, 2020 at 0:17 | history | protected | VE7JRO | ||
| Feb 6, 2020 at 0:08 | answer | added | acicuc | timeline score: 2 | |
| Mar 18, 2019 at 16:29 | answer | added | Wagner Lip | timeline score: 3 | |
| Oct 29, 2017 at 21:50 | answer | added | Arnadath | timeline score: 2 | |
| Jun 3, 2017 at 16:47 | comment | added | Transistor Overlord | @EdgarBonet Sorry for the late reply. I dont have any current quantified fixed accuracy. I just want to know all possible options for now | |
| Apr 30, 2017 at 7:36 | comment | added | Edgar Bonet | Could you quantify your accuracy requirement? The approximation cos(π/2x) ≈ 1−x² has a maximal error of 5.6e-2. And (1−x²)(1−0.224x²), which costs 3 multiplications, is good to within 9.20e-4. | |
| Apr 29, 2017 at 20:16 | history | tweeted | twitter.com/StackArduino/status/858414624114343936 | ||
| Apr 29, 2017 at 19:21 | answer | added | Edgar Bonet | timeline score: 7 | |
| Apr 29, 2017 at 18:07 | answer | added | Halzephron | timeline score: 8 | |
| Apr 29, 2017 at 12:25 | answer | added | JRobert | timeline score: 3 | |
| Apr 29, 2017 at 11:49 | answer | added | sa_leinad | timeline score: 4 | |
| Apr 29, 2017 at 11:44 | answer | added | sa_leinad | timeline score: 4 | |
| Apr 29, 2017 at 11:34 | answer | added | Majenko | timeline score: 11 | |
| Apr 29, 2017 at 11:34 | answer | added | dannyf | timeline score: 1 | |
| Apr 29, 2017 at 11:13 | comment | added | Majenko | For just 90 (integer) degrees a 90-entry lookup table would be fastest and most efficient. In fact for the full 360 degrees you can use a 90-entry lookup table. Just read it backwards for 90-179 and invert it for 180-269. Do both for 270-359. | |
| Apr 29, 2017 at 10:59 | comment | added | Transistor Overlord | Degrees yes. I think it would be easier to write code and test if we use integers, so I would go with that. I ll put more clear information about edits | |
| Apr 29, 2017 at 10:26 | comment | added | sa_leinad | I assume you are wanting to work in degrees. Are you going to want to input integers or decimal numbers for the angle? | |
| Apr 29, 2017 at 10:09 | history | edited | Transistor Overlord | CC BY-SA 3.0 |
Mentioned tolerance for error and approximation
|
| S Apr 29, 2017 at 10:07 | history | suggested | Ghanima | CC BY-SA 3.0 |
fixed indentation, minor copy editing
|
| Apr 29, 2017 at 10:07 | comment | added | Transistor Overlord | Yes actually, but I would like to know the extent of the error of different methods. This is not a precision product but a side project of mine. Actually approximations are inevitable for almost any (if not any) digital system implementing a mathematical function | |
| Apr 29, 2017 at 10:03 | comment | added | sa_leinad | Would you be ok with approximate values? | |
| Apr 29, 2017 at 9:44 | review | Suggested edits | |||
| S Apr 29, 2017 at 10:07 | |||||
| Apr 29, 2017 at 9:11 | review | First posts | |||
| Apr 29, 2017 at 10:09 | |||||
| Apr 29, 2017 at 9:09 | history | asked | Transistor Overlord | CC BY-SA 3.0 |