Package park :: Package fitting :: Module rangemap :: Class Asymptote

Class Asymptote

source code

Logarithmic asymptote function.

This is an asymptote function which follows the floating point representation, using the following mapping:

[-1e308,-1e-308] -> [0,0.5)
0 -> 0.5
[1e-308,1e308] -> (0.5+eps,1]

With 11 bits of the 53 bits of available precision for the exponent, this leaves a tolerance of 1e-12 on the fitting parameters, which is easily good enough for any real unbounded fitting problem. The user can increased the precision on the bounded parameters by using tighter bounds.

Instance Methods
 
forward(x)
Linearize floating point values using an exponential scale.
source code
 
inverse(v)
Restore floating point value from linear form on an exponential scale.
source code
Class Variables
  scale = 0
  offset = -2047
Method Details

forward(x)

source code 

Linearize floating point values using an exponential scale.

Convert sign*m*2^e to sign*(e+1023+m), yielding a value in [-2047,2047]

inverse(v)

source code 

Restore floating point value from linear form on an exponential scale.

Convert sign*(e+1023+m) to sign*m*2^e, yielding a value in [-1e308,1e308]