TMathControl V1.0 Component for Borland Delphi®
Copyright ©1996, John Merrit, Analog Species in a Digital World Productions.
e-mail: john.merrit@NET4ALL.BE
All Rights Reserved.


Built in functions and properties of TMathControl:
*Trigonometry: sin, cos, tan, arcsin, arccos, arctan;
*Hyperbolic: sinh, cosh, tanh, arcsinh, arccosh, arctanh;
*Logaritmic: log, ln, alog, exp;
*power: x to power y, y-th root of x, 1/x, x², sqrt(x);
*Misc. complex number algebra: absolute value, argument, conjugate, 
 a real number -> complex number and vice versa; 
*Statistic: combinations, permutations, faculty (windows calculator =< 170, mine > 1500!)
    
These functions work on both real and complex numbers (if appropriate).
To reduce errormessages, some functions allow input of a complex number while they actually require a
real number. The complex part will be omitted in this cases.
e.g.
Faculty requires a real (actually a positive integer*) number.
If you entered the complex number (25,12) and push the faculty button (!), the calculator
calculates 25! instead of returning an errormessage.
This is the case for the following functions:
rad to degrees (R->D), degrees to rad (D->R), faculty (!), combinations (Comb), permutations (Perm).
Also, real numbers will be converted to integer (round) when needed
i.e.
  3.25! will calculate 3!.
(!, combinations (Comb), permutations (Perm))

*To make bigger (smaller) calculations possible, a trick has being used to force these functions to
 work with (the integer part) of a real (extended) number.
 For example:
 Using longint as datatype would restrict us to calculate faculties for numbers =< 12.
 Now, the control is capable of calculating the faculty of a number > 1500.
 I've never needed this BIG numbers but knowing how for example Newton's Binomium works, it's
 easy to believe someone has some benefits from them. 

Controlproperties:
*Accuracy: accuracy of calculations (optional, see source code);
*MathFunc: function (sin, cos, etc...);


Controlevents (speak for themself):
Note that the control uses extended datatypes (3.4 * 10-4932 ... 1.1 * 10+4932) so you'll have to
do strange things if you want (for example) to make an overflow occur.
This has also one drawback: this control (and demoprogram) can't be used on a machine without a mathematical
coprocessor, but since we live in 1996, we'll assume that everybody has one...
*OnInvalidOp
*OnOverflow
*OnUnderflow  
*OnZeroDivide