Relative rounding - significant digits 15.
Mop-Up - Error limit 1E-15.
Note that the function mopup is used overall for improving the readability. The cells having
values greater than the limit are not modified.
(*) converts a well formed Excel formula to the equivalent in terms of Xnumber
multiprecision functions (xadd, xmult, etc.).
The advantage over x-evaluate is that the code can be debugged in normal excel using small
and in a familiar form. When the spreadsheet works it is ' converted to using nested x-calls for
the precision work. The intention is to tag the conversion onto the copy worksheet function ' so
that you end up with a multi-precision copy of the original.
Its main features are:
• converts a range of cells directly on site or in a new worksheet.
• skips cells having functions not convertible.
• skips array functions except MINVERSE and MMULT that are substituted with the
correspondent x-functions xMatInv and xMatMult.
• Errors (if any ) are shown on the panel
The digits parameter can be set in 4 different ways:
1) an integer number i.e. 30
2) a fixed reference of a cell, i.e. $A$1
3) a name of a cell, i.e. "digits"
4) simply nothing. In that case all the x-functions use the internal default = 30
Of course not all the Excel Functions can be converted. The list of the Excel functions
* , / , + , - , ^ , ABS , ACOS , ACOSH , ASIN , ASINH , ATAN , ATANH , AVERAGE
, COMBIN , COS , COSH , EXP , FACT , INT , LN , LOG , MDETERM , MINVERSE ,
MMULT , MOD , PI , PRODUCT , ROUND , SIN , SINH , SQRT , STDEV , STDEVP , SUM
, TAN , TANH , TRUNC , VAR , VARP
The conversion engine of this macro was originally developed by John Jones in an smart m4 macro language which
makes the VB native j-code more simple (!) to write. It appears in this package thanks to his courtesy
The functions are indicated with their original English names, that usually are different from the local names.