Comparing floats
Locked
 
|
Hi
I'm writing a simple converter between Celcius and Farhenheit as an example to start programming. Converter >> convertFarhenheit: anInteger ^ ((anInteger - 32) / 1.8) ConverterTest >> testFToC | converter | converter := TemperatureConverter new. self assert: ((converter convertFarhenheit: 86) = 30.0). self assert: ((converter convertFarhenheit: 50) = 10). self assert: ((converter convertFarhenheit: 52) = 11.11111111111111) My problem is with ((converter convertFarhenheit: 52) = 11.11111111111111) I do not want to have such ugly test. I tried either to control the output Converter >> convertFarhenheit: anInteger ^ ((anInteger - 32) / 1.8) roundDownTo: 0.1 Not a good idea Or to compare with closeTo: self assert: ((converter convertFarhenheit: 52) closeTo: 11.11) does not work because we cannot set the precision. So at ESUG I briefly discuss that with Nicolas Cellier and I need help. Should I introduce in pharo closeTo:interval: Any suggestions that a newby can understand is welcome. Stef |
|
|
Why not use "equalsTo:"?
Floats are tricky, you either round your result before comparing or change the comparison method. 0.333333333333333333 equalsTo: (1/3) "true" Regards! Esteban A. Maringolo 2016-09-02 15:41 GMT-03:00 stepharo <[hidden email]>: > Hi > > I'm writing a simple converter between Celcius and Farhenheit as an example > to start programming. > > > Converter >> convertFarhenheit: anInteger > ^ ((anInteger - 32) / 1.8) > > ConverterTest >> testFToC > > | converter | > converter := TemperatureConverter new. > self assert: ((converter convertFarhenheit: 86) = 30.0). > self assert: ((converter convertFarhenheit: 50) = 10). > self assert: ((converter convertFarhenheit: 52) = 11.11111111111111) > > > My problem is with ((converter convertFarhenheit: 52) = 11.11111111111111) > > I do not want to have such ugly test. > > I tried either to control the output > > Converter >> convertFarhenheit: anInteger > ^ ((anInteger - 32) / 1.8) roundDownTo: 0.1 > > Not a good idea > > Or to compare with closeTo: > > self assert: ((converter convertFarhenheit: 52) closeTo: 11.11) > > does not work because we cannot set the precision. > > So at ESUG I briefly discuss that with Nicolas Cellier and I need help. > > Should I introduce in pharo closeTo:interval: > > Any suggestions that a newby can understand is welcome. > > > Stef > > > |
|
|
In reply to this post by stepharo
You could implement a custom assert in ConverterTest to do the comparison, something like
assert: aNumber withPrecision: precision equals: otherNumber self assert: (aNumber round: precision) equals: otherNumber assert: aNumber closeTo: otherNumber assert: aNumber withPrecision: self defaultPrecision equals: otherNumber defaultPrecision ^ 2 Best regards, Henrik -----Original Message----- From: Pharo-users [mailto:[hidden email]] On Behalf Of stepharo Sent: Friday, September 2, 2016 8:41 PM To: Any question about pharo is welcome <[hidden email]> Subject: [Pharo-users] Comparing floats Hi I'm writing a simple converter between Celcius and Farhenheit as an example to start programming. Converter >> convertFarhenheit: anInteger ^ ((anInteger - 32) / 1.8) ConverterTest >> testFToC | converter | converter := TemperatureConverter new. self assert: ((converter convertFarhenheit: 86) = 30.0). self assert: ((converter convertFarhenheit: 50) = 10). self assert: ((converter convertFarhenheit: 52) = 11.11111111111111) My problem is with ((converter convertFarhenheit: 52) = 11.11111111111111) I do not want to have such ugly test. I tried either to control the output Converter >> convertFarhenheit: anInteger ^ ((anInteger - 32) / 1.8) roundDownTo: 0.1 Not a good idea Or to compare with closeTo: self assert: ((converter convertFarhenheit: 52) closeTo: 11.11) does not work because we cannot set the precision. So at ESUG I briefly discuss that with Nicolas Cellier and I need help. Should I introduce in pharo closeTo:interval: Any suggestions that a newby can understand is welcome. Stef |
|
|
how about having assert:closeTo: ?
> On 02 Sep 2016, at 21:09, Henrik Nergaard <[hidden email]> wrote: > > You could implement a custom assert in ConverterTest to do the comparison, something like > > assert: aNumber withPrecision: precision equals: otherNumber > > self > assert: (aNumber round: precision) > equals: otherNumber > > assert: aNumber closeTo: otherNumber > > assert: aNumber withPrecision: self defaultPrecision equals: otherNumber > > defaultPrecision > ^ 2 > > > Best regards, > Henrik > > -----Original Message----- > From: Pharo-users [mailto:[hidden email]] On Behalf Of stepharo > Sent: Friday, September 2, 2016 8:41 PM > To: Any question about pharo is welcome <[hidden email]> > Subject: [Pharo-users] Comparing floats > > Hi > > I'm writing a simple converter between Celcius and Farhenheit as an example to start programming. > > > Converter >> convertFarhenheit: anInteger > ^ ((anInteger - 32) / 1.8) > > ConverterTest >> testFToC > > | converter | > converter := TemperatureConverter new. > self assert: ((converter convertFarhenheit: 86) = 30.0). > self assert: ((converter convertFarhenheit: 50) = 10). > self assert: ((converter convertFarhenheit: 52) = 11.11111111111111) > > > My problem is with ((converter convertFarhenheit: 52) = 11.11111111111111) > > I do not want to have such ugly test. > > I tried either to control the output > > Converter >> convertFarhenheit: anInteger > ^ ((anInteger - 32) / 1.8) roundDownTo: 0.1 > > Not a good idea > > Or to compare with closeTo: > > self assert: ((converter convertFarhenheit: 52) closeTo: 11.11) > > does not work because we cannot set the precision. > > So at ESUG I briefly discuss that with Nicolas Cellier and I need help. > > Should I introduce in pharo closeTo:interval: > > Any suggestions that a newby can understand is welcome. > > > Stef > > > |
|
|
would be nice to extend Sunit with that
Le 3/9/16 à 00:21, Yuriy Tymchuk a écrit : > how about having assert:closeTo: ? > >> On 02 Sep 2016, at 21:09, Henrik Nergaard <[hidden email]> wrote: >> >> You could implement a custom assert in ConverterTest to do the comparison, something like >> >> assert: aNumber withPrecision: precision equals: otherNumber >> >> self >> assert: (aNumber round: precision) >> equals: otherNumber >> >> assert: aNumber closeTo: otherNumber >> >> assert: aNumber withPrecision: self defaultPrecision equals: otherNumber >> >> defaultPrecision >> ^ 2 >> >> >> Best regards, >> Henrik >> >> -----Original Message----- >> From: Pharo-users [mailto:[hidden email]] On Behalf Of stepharo >> Sent: Friday, September 2, 2016 8:41 PM >> To: Any question about pharo is welcome <[hidden email]> >> Subject: [Pharo-users] Comparing floats >> >> Hi >> >> I'm writing a simple converter between Celcius and Farhenheit as an example to start programming. >> >> >> Converter >> convertFarhenheit: anInteger >> ^ ((anInteger - 32) / 1.8) >> >> ConverterTest >> testFToC >> >> | converter | >> converter := TemperatureConverter new. >> self assert: ((converter convertFarhenheit: 86) = 30.0). >> self assert: ((converter convertFarhenheit: 50) = 10). >> self assert: ((converter convertFarhenheit: 52) = 11.11111111111111) >> >> >> My problem is with ((converter convertFarhenheit: 52) = 11.11111111111111) >> >> I do not want to have such ugly test. >> >> I tried either to control the output >> >> Converter >> convertFarhenheit: anInteger >> ^ ((anInteger - 32) / 1.8) roundDownTo: 0.1 >> >> Not a good idea >> >> Or to compare with closeTo: >> >> self assert: ((converter convertFarhenheit: 52) closeTo: 11.11) >> >> does not work because we cannot set the precision. >> >> So at ESUG I briefly discuss that with Nicolas Cellier and I need help. >> >> Should I introduce in pharo closeTo:interval: >> >> Any suggestions that a newby can understand is welcome. >> >> >> Stef >> >> >> > > |
|
|
In reply to this post by Henrik Nergaard
Thanks I will try that.
Stef Le 2/9/16 à 21:09, Henrik Nergaard a écrit : > You could implement a custom assert in ConverterTest to do the comparison, something like > > assert: aNumber withPrecision: precision equals: otherNumber > > self > assert: (aNumber round: precision) > equals: otherNumber > > assert: aNumber closeTo: otherNumber > > assert: aNumber withPrecision: self defaultPrecision equals: otherNumber > > defaultPrecision > ^ 2 > > > Best regards, > Henrik > > -----Original Message----- > From: Pharo-users [mailto:[hidden email]] On Behalf Of stepharo > Sent: Friday, September 2, 2016 8:41 PM > To: Any question about pharo is welcome <[hidden email]> > Subject: [Pharo-users] Comparing floats > > Hi > > I'm writing a simple converter between Celcius and Farhenheit as an example to start programming. > > > Converter >> convertFarhenheit: anInteger > ^ ((anInteger - 32) / 1.8) > > ConverterTest >> testFToC > > | converter | > converter := TemperatureConverter new. > self assert: ((converter convertFarhenheit: 86) = 30.0). > self assert: ((converter convertFarhenheit: 50) = 10). > self assert: ((converter convertFarhenheit: 52) = 11.11111111111111) > > > My problem is with ((converter convertFarhenheit: 52) = 11.11111111111111) > > I do not want to have such ugly test. > > I tried either to control the output > > Converter >> convertFarhenheit: anInteger > ^ ((anInteger - 32) / 1.8) roundDownTo: 0.1 > > Not a good idea > > Or to compare with closeTo: > > self assert: ((converter convertFarhenheit: 52) closeTo: 11.11) > > does not work because we cannot set the precision. > > So at ESUG I briefly discuss that with Nicolas Cellier and I need help. > > Should I introduce in pharo closeTo:interval: > > Any suggestions that a newby can understand is welcome. > > > Stef > > > |
|
|
In reply to this post by Uko2
JUnit has the equivalent of: assert:equals:range:
It could then be: self assert: 11.11 equals: 11.11 range: 0.0001 Alexandre > On Sep 2, 2016, at 7:21 PM, Yuriy Tymchuk <[hidden email]> wrote: > > how about having assert:closeTo: ? > >> On 02 Sep 2016, at 21:09, Henrik Nergaard <[hidden email]> wrote: >> >> You could implement a custom assert in ConverterTest to do the comparison, something like >> >> assert: aNumber withPrecision: precision equals: otherNumber >> >> self >> assert: (aNumber round: precision) >> equals: otherNumber >> >> assert: aNumber closeTo: otherNumber >> >> assert: aNumber withPrecision: self defaultPrecision equals: otherNumber >> >> defaultPrecision >> ^ 2 >> >> >> Best regards, >> Henrik >> >> -----Original Message----- >> From: Pharo-users [mailto:[hidden email]] On Behalf Of stepharo >> Sent: Friday, September 2, 2016 8:41 PM >> To: Any question about pharo is welcome <[hidden email]> >> Subject: [Pharo-users] Comparing floats >> >> Hi >> >> I'm writing a simple converter between Celcius and Farhenheit as an example to start programming. >> >> >> Converter >> convertFarhenheit: anInteger >> ^ ((anInteger - 32) / 1.8) >> >> ConverterTest >> testFToC >> >> | converter | >> converter := TemperatureConverter new. >> self assert: ((converter convertFarhenheit: 86) = 30.0). >> self assert: ((converter convertFarhenheit: 50) = 10). >> self assert: ((converter convertFarhenheit: 52) = 11.11111111111111) >> >> >> My problem is with ((converter convertFarhenheit: 52) = 11.11111111111111) >> >> I do not want to have such ugly test. >> >> I tried either to control the output >> >> Converter >> convertFarhenheit: anInteger >> ^ ((anInteger - 32) / 1.8) roundDownTo: 0.1 >> >> Not a good idea >> >> Or to compare with closeTo: >> >> self assert: ((converter convertFarhenheit: 52) closeTo: 11.11) >> >> does not work because we cannot set the precision. >> >> So at ESUG I briefly discuss that with Nicolas Cellier and I need help. >> >> Should I introduce in pharo closeTo:interval: >> >> Any suggestions that a newby can understand is welcome. >> >> >> Stef >> >> >> > > -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. |
«
Return to Pharo Smalltalk Users
|
1 view|%1 views
| Free forum by Nabble | Edit this page |