1.5.0

• sp_sph_harm has been removed for now as it's deprecated in scipy
• Number Types and Unit Systems are actual objects, rather than strings
• Additional input number type checks (e.g. positivity, integers) for operations that need them
• Raise a NumberTooBigException when trying to render a number that is too big for the maximum Number size/prec
• Fixed the number type of an operation never being decimal, even when the first parameter is decimal
• Fixed setting the number of decimal places before decimal point, correctly sets the total precision so we don't have more decimal places than total size

1.4.5

• Fixed crash for operation exceptions with non-decimal numbers

1.4.4

• Fixed 'as' operator dropping value's unit
• exp function has increased precision
• fact, lcd, gcd functions and << and >> operators correctly check input for negativity/non-integers

1.4.3

• Fixed log function behaviour when base is non-integer, zero, or negative
• Refactored operator/function error messages to make them more helpful and consistent
• Fixed crash when fetching an element from an array that is an exception

1.4.2

• Fixed trigonometry functions not returning Number instances

1.4.1

• Fixed floor and ceil functions when passing the places parameter
• Fixed round, floor and ceil functions with percentage numbers
• Fixed boolean conditional operators returning value with unit, when inputs have units

1.4.0

• New functions: lcm (lowest common multiple) and gcd (greatest common divisor).
• Added many special functions from SciPy.
• Added 'as' operator that can convert a value's numerical type (e.g. convert to binary).
• Added percentage numerical type.
• Added inline functions.
• Nested arrays are now possible.
• Function names can now contain, but not start with, a number.
• Fixed a unit assignment not having higher precedence than multiply.

1.3.1

• Fixed displaying units - divisor in brackets now used closed bracket properly.
• Fixed potential infinite loop when simplifying units. This has slightly reduced the kind of simplification that can be done.
• Unit powers can now be non-integers.
• Assignment operators (e.g. +=) now return the new value of the variable.
• Number power operator is now more accurate when given negative powers.

1.3.0

• Ceil and floor functions now also have an optional places parameter.
• Numerical Engine has number_set_rounding function to set rounding mode, use names of decimal rounding modes.
• Numerical Engine has number_size_set function to set the maximum size of numbers (before decimal point).
• Number casters now includes a reference to the function to reverse the casting to Number.
• Various improvements for alternative-base numbers and scientific notation numbers:
  • Converting a number to an alternative-base number, or scientific notation number, respects the calculator's precision option.
  • Alternative-base numbers and scientific notation numbers now all stored internally as a Number, with 'number_cast' attribute storing a function reference to convert back to its original representation - this is done when casting to a string. This avoids converting between formats multiple times internally, potentially losing precision each time.
  • Fixing scientific notation number 0E0 being displayed as E0.
  • Round, ceil and floor functions now all work properly with alternative-base numbers when specifying places.
  • Binary numbers now preserve their width (e.g. number of leading zeros) on creation and after a bitwise operation, all other operations discard the width afterwards.
• Fixing error thrown when throwing a CalculateException.

1.2.2

• Bug fix: Closing file handle in externalfunctions.py.