Hi,
Is there a easy way to cast BigNumber to double or float? or the otherway around?
Thank
Per
Hi,
Is there a easy way to cast BigNumber to double or float? or the otherway around?
Thank
Per
Have you tried something like this?
#include <BigNumber.h>
float casted;
BigNumber big;
void setup(){
BigNumber::begin();
}
void loop(){
casted = (float) big;
}
I am trying to use BigNumber.h for big number variables and for doing calculations with them. so far i am getting 0 as result. Am i doing something wrong?
#include <BigNumber.h>
#define MicrosMins 60000000 // micros to minutes multiplier
BigNumber currentPersent; // decimal must have 15+ resolution since will be like 0.0000000000000000002% on each look
BigNumber totalPersent; // decimal must have 15+ resolution
BigNumber lastPersent; // decimal must have 15+ resolution
BigNumber TimeMicros; // converting millis to micros will give a BIG number
unsigned long Time; // will store minuetes and will be variable acording to sensor input
unsigned long currentMathMicros;
unsigned long rotationMathMicros;
unsigned long previousMathMicros;
void setup()
{
BigNumber::begin (50);
Serial.begin(9600);
Time = 0;
rotationMathMicros = 0;
currentPersent = 0;
previousMathMicros = 0;
totalPersent = 0;
}
void loop()
{
currentMathMicros = micros ();
TimeMicros = Time * MicrosMins;
rotationMathMicros = currentMathMicros - previousMathMicros;
currentPersent = (100 * rotationMathMicros) / TimeMicros;
totalPersent = totalPersent + currentPersent;
previousMathMicros = currentMathMicros;
Serial.print("BIG NUM Persent Done: "); Serial.print (totalPersent); Serial.println(" %");
Serial.print("BUG NUM Time Micros: "); Serial.println (TimeMicros);
}
DonPedro34:
Is there a easy way to cast BigNumber to double or float? or the otherway around?
You can convert a BigNumber to a double or float by using the normal libraries (on the Uno double and float are the same anyway). For example:
#include <BigNumber.h>
void setup()
{
BigNumber::begin (50);
Serial.begin(115200);
Serial.println ();
BigNumber foo ("123.456789");
char * s = foo.toString ();
float bar = atof (s);
free (s);
Serial.print ("foo = ");
Serial.println (foo);
Serial.print ("bar = ");
Serial.println (bar, 8);
}
void loop()
{
}
Output:
foo = 123.456789
bar = 123.45678710
You can convert from a float to a BigNumber in a similar way by turning the float into a string, and then using that in a BigNumber contructor (like above). Note that floats don't have more than about 7 digits of precision.
Constantine86:
Am i doing something wrong?
Yes, you are converting to a BigNumber too late. See, for example: Integer and arithmetic overflow.
Also I don't see where Time is incremented:
TimeMicros = Time * MicrosMins;
Yes, you are converting to a BigNumber too late. See, for example: Integer and arithmetic overflow.
Also I don't see where Time is incremented:
Thank you for your fast response, as i am a newbe this casting trick looks very useful, does it work for doing maths with different type of variables? (int*float-long etc)
Time is constantly changing over each program loop based on sensor input and by doing another equation that outputs data to Time variable.
By to late you mean i should use BigNumber only at void setup? I thought i can use BigNumber as variables to store data and do some basic maths (+-*/) with them on void loop.
No I don't mean that.
#define MicrosMins 60000000 // micros to minutes multiplier
...
unsigned long Time;
So, "Time * MicrosMins" will be unsigned long. Which will probably overflow. Then you try to store that result into a BigNumber.
More like:
TimeMicros = BigNumber (Time) * BigNumber (MicrosMins);
Time is constantly changing over each program loop ...
Not in the code you posted, it isn't.
i added the Time variable equation and their variables.
#include <BigNumber.h>
#define MicrosMins 60000000 // micros to minutes multiplier
#define type 200
BigNumber currentPersent; // decimal must have 15+ resolution since will be like 0.00...002% on each loop
BigNumber totalPersent; // decimal must have 15+ resolution
BigNumber lastPersent; // decimal must have 15+ resolution
unsigned long TimeMicros; // converting millis to micros will give a BIG number
unsigned long Time; // will store minuetes and will be variable acording to sensor input
unsigned long currentMathMicros;
unsigned long rotationMathMicros;
unsigned long previousMathMicros;
unsigned int P; //user multiplier input multiplier from 1 to 100
float SensorInput; //sensor input
float castedPersent;
void setup()
{
BigNumber::begin (50);
Serial.begin(9600);
Time = 0;
rotationMathMicros = 0;
currentPersent = 0;
previousMathMicros = 0;
totalPersent = 0;
}
void loop()
{
Time = ((Type - (Type * (castedPersent / 100))) / SensorInput) * P;
currentMathMicros = micros ();
TimeMicros = BigNumber (Time) * BigNumber (MicrosMins);
rotationMathMicros = currentMathMicros - previousMathMicros;
currentPersent = (100 * rotationMathMicros) / TimeMicros;
totalPersent = totalPersent + currentPersent;
previousMathMicros = currentMathMicros;
char * s = totalPersent.toString ();
float castedPersent = atof (s);
free (s);
Serial.print("BIG NUM Persent Done: "); Serial.print (totalPersent); Serial.println(" %");
Serial.print("BUG NUM Time Micros: "); Serial.println (TimeMicros);
}
TimeMicros outputs 4.294.781.440 while it should output 360.000.000
Can you give an example of the right BigNumber syntax for doing some basic maths over on void loop?
BigNumber::begin(50); 50 gives the digit resolution, right? Gammon Forum : Electronics : Microprocessors : Arbitrary precision (big number) library port for Arduino i am trying to see if i can find more info on this post
Time = ((Type - (Type * (castedPersent / 100))) / SensorInput) * P;
Error compiling:
/tmp/arduino_modified_sketch_725756/sketch_aug24a.ino: In function 'void loop()':
sketch_aug24a:37: error: 'Type' was not declared in this scope
Time = ((Type - (Type * (castedPersent / 100))) / SensorInput) * P;
This is a bit time-wasting, posting code that doesn't compile and asking why it outputs certain results.
This thread is really nothing to do with the announcement thread about the library. I have split it off into a new thread.