Converting String to floating number using toFloat()

Using Arduino Programming Questions
1 
void setup() {
  Serial.begin(9600);
}

void loop() 
{
 String inputString;
 inputString.reserve(100);
 inputString = "01:36:59  $GNGGA,013659.00,3402.77961,N,11750.73009,W,1,10,1.87,222.2,M,-32.3,M,,*7C";
 inputString.remove(12,1);
int index1 = inputString.indexOf(",N"); // Find the index of "N"
int index2 = inputString.indexOf(",W"); // Find the index of "W"

// Extract the latitude value and convert it to a float
String latString = inputString.substring(index1-10, index1);
Serial.println(latString);
float latitude = latString.toFloat();
// Extract the longitude value and convert it to a float
String lonString = inputString.substring(index2-11, index2);
float longitude = lonString.toFloat();
Serial.println(lonString);

// Print the latitude and longitude values
Serial.print("Latitude:");
Serial.println(latitude, 5);
Serial.print("Longitude:");
Serial.println(longitude, 5);
}

In theory this code should parse through the string and extract the values before the N and W which are latitude and longitude values. However when converting the string to a float using toFloat() it changes the last two decimals places. I checked to see if I was parsing through the string correctly by printing the strings individually, so I'm kinda lost as to why the value changes when it gets converted.

2

Because floats have only 5-6 significant digits.

3

Is there a way to have more significant digits?

4

On 32 bit processors like the ESP32 you could use double.

5

Is there any other way to solve it within the arduino IDE?

6

It is not a matter of the IDE, ESP32 are supported.

You could parse the integral and the fractional part into separate variables.

7

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.