GY-63_MS5611/libraries/FLE
2023-01-23 10:10:49 +01:00
..
.github 0.1.3 FLE + github actions 2023-01-23 10:10:49 +01:00
examples 0.1.1 FLE 2021-05-28 13:27:47 +02:00
test 0.1.3 FLE 2022-12-08 17:29:26 +01:00
.arduino-ci.yml 0.1.3 FLE 2022-12-08 17:29:26 +01:00
CHANGELOG.md 0.1.3 FLE 2022-12-08 17:29:26 +01:00
FLE.cpp 0.1.3 FLE 2022-12-08 17:29:26 +01:00
FLE.h 0.1.3 FLE 2022-12-08 17:29:26 +01:00
keywords.txt 2021-01-29 2021-01-29 12:31:58 +01:00
library.json 0.1.3 FLE 2022-12-08 17:29:26 +01:00
library.properties 0.1.3 FLE 2022-12-08 17:29:26 +01:00
LICENSE 0.1.3 FLE + github actions 2023-01-23 10:10:49 +01:00
README.md 0.1.3 FLE 2022-12-08 17:29:26 +01:00

Arduino CI Arduino-lint JSON check License: MIT GitHub release

FLE

Arduino library for Arduino library for float with error datatype

Description

This experimental library provides basic math when have a quantity with a certain error margin. E.g. when you approximate PI as 22/7 it is not exact. By doing the math with FLE's one can see how much error adds up in an calculation.

Interface

Printable

The FLE class implements the public interface of Printable. This allows you to print an FLE in human readable form.

    FLE x(3.14, 0.002);
    x.setDecimals(4);
    Serial.println(x);                 // will print 3.1400 ± 0.0020
    Serial.println(x.setDecimals(2));  // will print 3.14 ± 0.00
    Serial.println(x.value());         // will print 3.14

When the ± char does not print correctly, one could change the font.

Functions

  • FLE(val = 0, err = 0) constructor, with default value and error set to 0.
  • PrintTo(Print& p) printable interface, define a stream to print to.
    Format is "value ± error" see above (plus-minus ± = char(0177))
  • setDecimals(n) will print the FLE both value and error with n decimals.
  • setSeparator(char c) overrules the standard ± char. (0177)
  • value() returns value part.
  • error() return error part.
  • relError() returns relative error, except when value == 0. Then the function returns 0. Q: should this be "NaN" ?
  • high() returns value + error margin (= max real value).
  • low() returns value - error margin (= min real value).

Furthermore the basic math is implemented, "+, -, *, /, +=, -=, *=, /="

Set like functions

  • bool in(FLE y) x.in(y) returns true if x lies completely in y (range is a subset)
  • FLE shared(FLE y) returns the overlapping range.

Weak propositions

Experimental.

  • bool peq(FLE &y) possible equal.
  • bool pne(FLE &y) possible not equal.
  • bool plt(FLE &y) possible less than.
  • bool ple(FLE &y) possible less equal.
  • bool pgt(FLE &y) possible greater than.
  • bool pge(FLE &y) possible greater equal.

Todo

  • update documentation
  • comparison (investigate, what means equal or less than ..)
  • functions log, exp,
  • functions sqr, sqrt, pow
  • functions sin, cos, tan (+ inverse + hyp)
  • test ad infinitum
  • option to set the ± char for platforms that cannot print it.
  • test negative numbers
  • test large / small numbers.
  • test other separator
  • more demo sketches...

Operation

  • negative numbers not tested yet

See example