Category : C Source Code
Archive   : CEPHES22.ZIP
Filename : CALC100.DOC

CALC100 version 1.2

CALC100 is a 100 digit calculator program. It can do
arithmetic, Fortran-style algebraic syntax, and various
built-in mathematical functions. Its fast floating
point arithmetic takes advantage of the computer's unsigned
integer multiply and divide instructions. Division uses
a Newton-Raphson iteration, which is much faster than
the shift-and-subtract approach.

Algebraic expressions are entered in the same syntax as
Fortran or other high level language statements. Multiple
statements can be entered on the same line but must be separated
by semicolons. Thus the entry 'x=2; y=x+4' sets the variables
x and y as indicated and also displays the numeric results.
There are 7 algebraic variables named t, u, v, w, x, y, and z.
The constant pi is also represented as a variable.

All of calc100's commands have the syntax of C language
functions. For example, to display the logarithm of 2, you may
enter the command 'log(2)' followed by a carriage return. The
parentheses may be omitted if there is no ambiguity; thus 'log 2'
gives the same result.

acos Radian inverse circular cosine
acosh Inverse hyperbolic cosine
asin Radian inverse circular sine
asinh Inverse hyperbolic sine
atan Radian inverse circular tangent
atanh Inverse hyperbolic tangent
bits Display floating point number as array of integers
cbrt Cube root
cos Circular radian cosine
cosh Hyperbolic cosine
cot Circular radian cotangent
digits N Set number of digits in the numeric display to N
dm Define macro command (see text below)
em N Execute macro N times
exit Terminate program - use at end of command files
exp Base e exponential function
fac Factorial of integer argument
gamma Gamma function (accuracy >= 55 decimals)
h or help Display list of commands
hex Display integer argument in radix 8, 10, and 16
jv(v,x) Bessel function Jv(x) (accuracy about 40 decimals)
log Natural logarithm
pow(x,y) exp( y log x ) = x raised to the yth power
save filename Save session in file 'filename'
or in 'calc.sav' if no file name given
sin Circular radian sine
sinh Hyperbolic sine
sqrt Square root
system command Execute operating system 'command' and return
take filename Read commands from indicated disc file
tan Circular radian tangent
tanh Hyperbolic tangent
tm Type (display) currently defined macro command
yn(n,x) Bessel function Yn(x) (accuracy about 40 decimals)

To define a macro expression, enter 'dm' and type the
expression on the next line. The expression must fit on
one line, but multiple statements can be entered if they
are separated by a semicolon (;). Execute the defined
macro N times by the command 'em N'. Display the current
macro definition by 'tm'. Example:

x=y=v=z=1
dm
z=z*x/v; v=v+1; y=y+z
em 10

is a power series for exp(1). You can put the above into
an ASCII file named, say, sample.tak and execute the program
from the file by the command 'take sample.tak'.

You can also redirect input or output, for example: 'calc
answer.dat'. But in that case there shoud be an
'exit' command at the end of the input file.

The 'system' command passes the rest of the command line
to the computer's operating system to execute as a command.
After the command has finished executing, control will return
to the calc100 program.

The 'save' command starts logging most of what appears
on the screen into a disc file. Certain error printouts and
the help menus will not be written into the file. A repeated
invocation of 'save' closes the old save file and opens a new one
(or possibly the same one if the file name argument is the same).
Printout generated by a 'system' command will not go into the
save file, but it probably will go into a file that was created by
redirecting the standard output as in 'calc answer.dat'.

The rigorous way of calling a function is to put the
argument in parentheses. However if no expression decoding
is implied, the parentheses can be omitted. For example
2 + sin x means 2 + sin(x), but sin x + 2 means sin(x) + 2.
If a function has more than one argument, the parentheses
must be there. For example pow(2,3) is 8, but pow 2,3 is
an error.

By default, the numeric display gives 70 digits after the
decimal point. This may be varied from N = 1 to 99 digits by
the command 'digits N'.

-- Steve Moshier
September, 1988