--[[ This file contains the number class which allows to do calculations with uncertainties. Copyright (c) 2021 Thomas Jenni Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. ]]-- -- Number class local Number = {} Number.__index = Number -- make the table callable setmetatable(Number, { __call = function(class, ...) return Number.new(...) end }) -- If a number is given as a string and it has no uncertainty defined, this options -- allows to set a default uncertainty, i.e. (5.6) will become (5.60 +/- 0.05) Number.defaultUncertainty = 0.5 -- Switch for writing the uncertainty or not Number.omitUncertainty = true -- If true, the plus-minus notation will be used, otherwise the uncertainty -- will be appended to the value in parentheses. Number.seperateUncertainty = true -- number format "decimal" or "scientific" Number.DECIMAL = 0 Number.SCIENTIFIC = 1 Number.format = Number.SCIENTIFIC -- The rounding function uses this value to determine the rounding method -- if it is set to 0.5 the rounding method is "towards nearest, ties to even". Number.half = 0.5 -- constructor function Number.new(x, dx) local n = {} setmetatable(n, Number) -- pure number without uncertainty if type(x) == "number" then n._x = x n._dx = dx or 0 return n -- copy constructor elseif type(x) == "table" and getmetatable(x) == Number then n._x = x._x n._dx = x._dx return n -- number from string elseif type(x) == "string" then -- parse string of the form "3.4e-3" local m = tonumber(x) if m ~= nil then n._x = m local m1, m2, e = string.match(x,"^([^%s%.]*)%.?([^%seE]*)[eE]?([^%s]*)$") -- get exponent if e ~= "" then e = tonumber(e) else e = 0 end -- calculate uncertainty n._dx = Number.defaultUncertainty * 10^e if m2 ~= "" then n._dx = n._dx * 10^(-string.len(m2)) end return n end -- parse string of the form "5.4e-3 +/- 2.4e-6" local m1, m2 = string.match(x,"^([^%s]*)%s*%+%/%-%s*([^%s]*)$") if m1 ~= nil and m2 ~= nil then n._x = tonumber(m1) n._dx = tonumber(m2) return n end -- parse string of the form "5.45(7)e-23" local m1, m2, u, e = string.match(x,"^([^%s%.]*)%.?([^%seE]*)%(([0-9]+)%)[eE]?([^%s]*)$") if m1 ~= nil then n._x = tonumber(m1.."."..m2) n._dx = u * 10^(-string.len(m2)) -- exponent if e ~= "" then n._x = n._x * 10^e n._dx = n._dx * 10^e end return n end print("Error: The string '"..tostring(_x).."' cannot be converted to a physical.Number.") return nil -- default else n._x = 0 n._dx = 0 return n end end -- return the mean value function Number:mean() return self._x end -- return the uncertainty function Number:uncertainty() return self._dx end -- get mantissa and exponent in scientific notation function Number._frexp(x) if x == 0 then return 0,0 end local s = (x < 0 and -1) or 1 local m = s * x local exp = math.floor(math.log(m,10)) m = s * m * 10^(-exp) return m, exp end -- round a number x to a precision n function Number._round(x,n) local m = math.pow(10.0, n) local y = x * m if y >= 0 then y = math.floor(y + Number.half) else y = math.ceil(y - Number.half) end return y / m end -- convert number to string -- x: number -- n: decimal places function Number._flt2str(x,n) return string.format("%."..math.max(n,0).."f", Number._round(x,n) ) end -- convert number to a number function Number:__tonumber() return self._x end -- plus minus notation, i.e. (5.040 +/- 0.001) function Number:toPlusMinusNotation(format, parenthesis, pmsign) if format == nil then format = Number.format end if parenthesis == nil then parenthesis = true end if pmsign == nil then pmsign = " +/- " end local m, e = self._frexp(self._x) local dm, de = self._frexp(self._dx) -- if the first digit of the uncetainty is a 1, give two digits of the uncertainty local udigit = 0 if math.floor(dm) == 1 then udigit = 1 end local str -- In the decimal format, the numbers are given as decimals, i.e. (0.02 +/- 0.001) if format == Number.DECIMAL then if de - udigit >= 0 then str = self._flt2str(self._x, 0)..pmsign..self._flt2str(self._dx, 0) else local digits = math.abs(-de + udigit) str = self._flt2str(self._x, digits)..pmsign..self._flt2str(self._dx, digits) end if parenthesis then str = "("..str..")" end -- In the scientific format, the numbers are written with powers of ten, i.e. (2.0 +/- 0.1) e-2 elseif format == Number.SCIENTIFIC then -- the uncertainty should have the same exponent as the value dm = dm * 10^(de-e) de = de - e if de >= 0 then str = self._flt2str(m, 0)..pmsign..self._flt2str(dm, 0) else local digits = math.abs(-de + udigit) str = self._flt2str(m, digits)..pmsign..self._flt2str(dm, digits) end if parenthesis then str = "("..str..")" end if e ~= 0 then str = str.."e"..e end else error("Unknown number format: '"..format.."'") end return str end -- generate a string representation in parenthesis notation, i.e. i.e. 5.45(7)e-23 -- Source: http://physics.nist.gov/cgi-bin/cuu/Info/Constants/definitions.html function Number:toParenthesisNotation(format) if format == nil then format = Number.format end local m, e = self._frexp(self._x) local dm, de = self._frexp(self._dx) -- if the first digit of the uncetainty is a 1, give two digits of the uncertainty local udigit = 0 if self._round(dm, 0) == 1 then udigit = 1 dm = dm * 10 de = de - 1 end local str -- In the decimal format, the numbers are given as decimals, i.e. 0.00343(12) if format == Number.DECIMAL then if de - udigit >= 0 then str = self._flt2str(self._x, 0).."(".. self._flt2str(dm*10^de, 0) ..")" else str = self._flt2str(self._x, math.abs(de)).."(".. self._flt2str(dm, 0) ..")" end -- In the scientific format, the numbers are written with powers of ten, i.e. 3.43(12)e-3 elseif format == Number.SCIENTIFIC then str = self._flt2str(m, math.abs(de-e)).."("..self._flt2str(dm, 0)..")" if e ~= 0 then str = str.."e"..e end else error("Unknown number format: '"..format.."'") end return str end function Number:toOmitUncertaintyNotation(format) local m, e = self._frexp(self._x) local dm, de = self._frexp(self._dx) local str = "" if Number.format == Number.DECIMAL then str = self._flt2str(self._x, math.abs(de + 1)) elseif Number.format == Number.SCIENTIFIC then str = self._flt2str(m, math.abs(de + 1 - e)) if e ~= 0 then str = str.."e"..e end else error("Unknown number format: '"..format.."'") end return str end -- convert number to a string function Number:__tostring() if self._dx == 0 then return tostring(self._x) elseif Number.omitUncertainty then return self:toOmitUncertaintyNotation() elseif Number.seperateUncertainty then return self:toPlusMinusNotation() else return self:toParenthesisNotation() end end -- convert number to a string function Number:tosiunitx() if self._dx == 0 then return tostring(self._x) elseif Number.omitUncertainty then return self:toOmitUncertaintyNotation() elseif Number.seperateUncertainty then return self:toPlusMinusNotation(Number.format, false, "+-") else return self:toParenthesisNotation() end end -- equal -- Two physical numbers are equal if they have the same value and uncertainty function Number.__eq(n1,n2) return n1._x == n2._x and n1._dx == n2._dx end -- less than function Number.__lt(n1,n2) if getmetatable(n1) ~= Number then return n1 < n2._x elseif getmetatable(n2) ~= Number then return n1._x < n2 else return n1._x < n2._x end end -- less equal function Number.__le(n1,n2) if getmetatable(n1) ~= Number then return n1 <= n2._x elseif getmetatable(n2) ~= Number then return n1._x <= n2 else return n1._x <= n2._x end end -- add two numbers function Number.__add(n1,n2) local m = Number.new() if type(n1) == "number" then m._x = n1 + n2._x m._dx = n2._dx elseif type(n2) == "number" then m._x = n1._x + n2 m._dx = n1._dx -- n1 an object elseif getmetatable(n1) ~= Number then return n1.__add(n1,n2) -- n2 an object elseif getmetatable(n2) ~= Number then return n2.__add(n1,n2) else m._x = n1._x + n2._x m._dx = (n1._dx^2 + n2._dx^2)^0.5 end return m end -- subtract two numbers function Number.__sub(n1,n2) local m = Number.new() if type(n1) == "number" then m._x = n1 - n2._x m._dx = n2._dx elseif type(n2) == "number" then m._x = n1._x - n2 m._dx = n1._dx -- n1 an object elseif getmetatable(n1) ~= Number then return n1.__sub(n1,n2) -- n2 an object elseif getmetatable(n2) ~= Number then return n2.__sub(n1,n2) else m._x = n1._x - n2._x m._dx = (n1._dx^2 + n2._dx^2)^0.5 end return m end -- unary minus function Number.__unm(n1) local m = Number.new() m._x = -n1._x m._dx = n1._dx return m end -- multiplication function Number.__mul(n1,n2) local m = Number.new() -- n1 float , n2 number if type(n1) == "number" then m._x = n1 * n2._x m._dx = n1 * n2._dx -- n1 number , o2 float elseif type(n2) == "number" then m._x = n1._x * n2 m._dx = n1._dx * n2 -- n1 is an object elseif getmetatable(n1) ~= Number then return n1.__mul(n1,n2) -- n2 is an object elseif getmetatable(n2) ~= Number then return n2.__mul(n1,n2) -- o1 number , o2 number else m._x = n1._x * n2._x local dx = 0 if n1._x ~= 0 then dx = (n1._dx/n1._x)^2 end if n2._x ~= 0 then dx = dx + (n2._dx/n2._x)^2 end m._dx = m._x * dx^0.5 end return m end -- division function Number.__div(n1,n2) local m = Number.new() -- n1 float , n2 number if type(n1) == "number" then m._x = n1 / n2._x m._dx = n2._dx * n1 / n2._x^2 -- n1 number , o2 float elseif type(n2) == "number" then m._x = n1._x / n2 m._dx = n1._dx / n2 -- n1 an object elseif getmetatable(n1) ~= Number then return n1.__div(n1,n2) -- n2 an object elseif getmetatable(n2) ~= Number then return n2.__div(n1,n2) -- o1 number , o2 number else m._x = n1._x / n2._x local dx = 0 if n1._x ~= 0 then dx = (n1._dx/n1._x)^2 end if n2._x ~= 0 then dx = dx + (n2._dx/n2._x)^2 end m._dx = m._x * dx^0.5 end return m end -- exponentiation function Number.__pow(n1,n2) local m = Number.new() -- n1 float , n2 number if type(n1) == "number" then m._x = n1^n2._x m._dx = n2._dx * math.abs(math.log(n1)*n1^n2._x) -- n1 number , o2 float elseif type(n2) == "number" then m._x = n1._x^n2 m._dx = n1._dx * math.abs(n2 * n1._x^(n2-1)) -- n1 an object elseif getmetatable(n1) ~= Number then return n1.__pow(n1,n2) -- n2 an object elseif getmetatable(n2) ~= Number then return n2.__pow(n1,n2) -- o1 number , o2 number else m._x = n1._x^n2._x m._dx = n1._dx*math.abs(n2._x*n1._x^(n2._x-1)) + n2._dx*math.abs(math.log(n1._x)*n1._x^n2._x) end return m end -- calculate the absolute value function Number.abs(n) local m = Number.new() m._x = math.abs(n._x) m._dx = n._dx return m end -- calculate the square root function Number.sqrt(n) local m = Number.new() m._x = math.sqrt(n._x) m._dx = n._dx / (2*math.sqrt(n._x)) return m end -- calculate the natural logarithm function Number.log(n,base) local m = Number.new() if base == nil then m._x = math.log(n._x) m._dx = n._dx / math.abs(n._x) else m._x = math.log(n._x,base) m._dx = n._dx / math.abs(n._x*math.log(base)) end return m end -- calculate the exponential function function Number.exp(n) local m = Number.new() m._x = math.exp(n._x) m._dx = n._dx * math.exp(n._x) return m end -- TRIGONOMETRIC FUNCTIONS -- https://en.wikipedia.org/wiki/Trigonometric_functions -- sine function Number.sin(n) local m = Number.new() m._x = math.sin(n._x) m._dx = n._dx * math.abs(math.cos(n._x)) return m end -- cosine function Number.cos(n) local m = Number.new() m._x = math.cos(n._x) m._dx = n._dx * math.abs(math.sin(n._x)) return m end -- tangent function Number.tan(n) local m = Number.new() m._x = math.tan(n._x) m._dx = n._dx * math.abs(1/math.cos(n._x)^2) return m end -- INVERS TRIGONOMETRIC FUNCTIONS -- https://en.wikipedia.org/wiki/Inverse_trigonometric_functions#arctan -- invers sine function Number.asin(n) local m = Number.new() m._x = math.asin(n._x) m._dx = n._dx * 1 / math.sqrt(1 - n._x^2) return m end -- inverse cosine function Number.acos(n) local m = Number.new() m._x = math.acos(n._x) m._dx = n._dx * 1 / math.sqrt(1 - n._x^2) return m end -- inverse tangent function Number.atan(n) local m = Number.new() m._x = math.atan(n._x) m._dx = n._dx / (1 + n._x^2) return m end -- HYPERBOLIC FUNCTIONS -- https://en.wikipedia.org/wiki/Hyperbolic_function -- hyperbolic sine function Number.sinh(n) local m = Number.new() m._x = 0.5 * math.exp(n._x) - 0.5 / math.exp(n._x) m._dx = n._dx * (0.5 * math.exp(n._x) + 0.5 / math.exp(n._x)) return m end -- hyperbolic cosine function Number.cosh(n) local m = Number.new() m._x = 0.5 * math.exp(n._x) + 0.5 / math.exp(n._x) m._dx = n._dx * (0.5 * math.exp(n._x) - 0.5 / math.exp(n._x)) return m end -- hyperbolic tangent function Number.tanh(n) local m = Number.new() m._x = (math.exp(n._x) - math.exp(-n._x)) / (math.exp(n._x) + math.exp(-n._x)) m._dx = n._dx / (0.5 * math.exp(n._x) + 0.5 / math.exp(n._x))^2 return m end -- INVERS HYPERBOLIC FUNCTIONS -- https://en.wikipedia.org/wiki/Inverse_hyperbolic_function -- inverse hyperbolic sine -- (-inf < n < +inf) function Number.asinh(n) local m = Number.new() m._x = math.log(n._x + math.sqrt(n._x^2 + 1)) m._dx = n._dx / math.sqrt(n._x^2 + 1) return m end -- inverse hyperbolic cosine -- (1 < n) function Number.acosh(n) local m = Number.new() m._x = math.log(n._x + math.sqrt(n._x^2 - 1)) m._dx = n._dx / math.sqrt(n._x^2 - 1) return m end -- inverse hyperbolic tangent -- (-1 < n < 1) function Number.atanh(n) local m = Number.new() m._x = 0.5 * math.log((1 + n._x)/(1 - n._x)) m._dx = n._dx / math.abs(n._x^2 - 1) return m end return Number