April 2014 - Pharo-users - lists.pharo.org

Re: [Pharo-users] Question on style

by Ben Coman

nacho wrote: > Camille, > > I now understood what you've said. It makes sense and I've changed it that > way. The only thing I'm not happy with is that implementing: > > QuadraticEquation>>calculateRoots > ^ self calculateRootsIfNone: [ self error: 'No real solution for the > equation' ] > > Makes the debugger appear and I would rather let the user know that there > are no real solutions. Bringing up the debugger can make one think that the > problem is in the code. > > What do you think of: > > QuadraticEquation>>calculateRoots > ^ self calculateRootsIfNone: [ ^ 'No real solution for the equation' ] > Thanks! > nacho > > > > > > ----- > Nacho > Smalltalker apprentice. > Buenos Aires, Argentina. > -- > View this message in context: http://forum.world.st/Question-on-style-tp4752165p4752388.html > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > > > How about... self inform: 'no real solutions' cheers -ben

April 2, 2014

Re: [Pharo-users] Drag and drop items between list views

by MartinW

First of all thank you for taking the time to explain all this to me! I already learned a lot about working with the spec widgets.. I still have a problem. The first version with the hardcoded collections did obviously change the original collections (that's the desired behaviour). Goubier Thierry wrote > list1 acceptDropBlock: > [ :transfer :event :source :receiver :index | > transfer passenger do: > [:e | > collection1 insert: e first before: index. > collection2 remove: e first]. > list1 items: collection1. > list2 items: collection2 ]. The last version though, did change the items in the list views, but not the original collections. Goubier Thierry wrote > list1 > acceptDropBlock: [ :transfer :event :source :receiver :index | > | c1 c2 l1 l2 | > l1 := source model model. > c1 := l1 listItems. > l2 := transfer source model model. > c2 := l2 listItems. > transfer passenger > do: [ :e | > c1 insert: e first before: index. > c2 remove: e first ]. > l1 items: c1. > l2 items: c2 ]. -- View this message in context: http://forum.world.st/Drag-and-drop-items-between-list-views-tp4752285p4752… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

April 2, 2014

Re: [Pharo-users] Question on style

by Alain Busser

Yes, and if there are no solution, you get an empty set instead of an error message. I find it more coherent, and simpler to use. On Wed, Apr 2, 2014 at 9:48 PM, nacho <0800nacho(a)gmail.com> wrote: > Alain, > Very good idea. A Set it's much better than an OrderedCollection. Already > changed that. Now when the solution is in both cases 0, it returns only (0) > and not (0 0). And also intuitively I agree with you that the solutions of > an equation are a set. > Thanks! > Nacho > > > > > ----- > Nacho > Smalltalker apprentice. > Buenos Aires, Argentina. > -- > View this message in context: > http://forum.world.st/Question-on-style-tp4752165p4752386.html > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > >

April 2, 2014

Re: [Pharo-users] Question on style

by nacho

Camille, I now understood what you've said. It makes sense and I've changed it that way. The only thing I'm not happy with is that implementing: QuadraticEquation>>calculateRoots ^ self calculateRootsIfNone: [ self error: 'No real solution for the equation' ] Makes the debugger appear and I would rather let the user know that there are no real solutions. Bringing up the debugger can make one think that the problem is in the code. What do you think of: QuadraticEquation>>calculateRoots ^ self calculateRootsIfNone: [ ^ 'No real solution for the equation' ] Thanks! nacho ----- Nacho Smalltalker apprentice. Buenos Aires, Argentina. -- View this message in context: http://forum.world.st/Question-on-style-tp4752165p4752388.html Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

April 2, 2014

Re: [Pharo-users] Question on style

by nacho

Alain, Very good idea. A Set it's much better than an OrderedCollection. Already changed that. Now when the solution is in both cases 0, it returns only (0) and not (0 0). And also intuitively I agree with you that the solutions of an equation are a set. Thanks! Nacho ----- Nacho Smalltalker apprentice. Buenos Aires, Argentina. -- View this message in context: http://forum.world.st/Question-on-style-tp4752165p4752386.html Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

April 2, 2014

Re: [Pharo-users] Question on style

by Camille Teruel

On 2 avr. 2014, at 18:45, nacho <0800nacho(a)gmail.com> wrote: > > Hi Nacho, > > I wouldn't throw an error because otherwise one must know beforehand that an > equation is non-resolvable or must use #on:do: everywhere. > Since that when you ask to resolve an equation you get back a collection of > solutions, a non-resolvable equation could just return an empty collection, > couldn't it?. > If you really want to consider that asking to resolve a non-resolvable > equation is an error, another solution would be to provide a #rootsIfNone: > method. That's a common pattern in Smalltalk. > I would also rename 'checkNegative' to 'discriminant', because it's not > obvious to see what that variable represents at first glance. > > Very good points. I came up with a third draft. This is evolving good! > Thanks > Nacho In fact, what I meant is that the function that compute the roots should be called #calculateRootIfNone: and take a block as parameter: QuadraticEquation>>calculateRootsIfNone: aBlock | discriminant | discriminant := linearCoefficient squared - ( 4 * quadraticCoefficient * constant). discriminant >= 0 ifFalse: aBlock ifTrue: [ ^ self solveRoots: discriminant] Like this, you can provide a #calcultaeRoots method like this: QuadraticEquation>>calculateRoots ^ self calculateRootsIfNone: [ self error: 'No real solution for the equation' ] And you could even provide a #calculateRootsSafely that just return an empty array. QuadraticEquation>>calculateRootsSafely ^ self calculateRootsIfNone: [ { } ] This is a common pattern in Smalltalk. Look at collection methods like #detect:ifNone: #at:ifPresent: #at:ifAbsent: ,etc ... BTW, is it expected that when the discriminant equals 0, the same root is returned twice? > > "The idea is to have a class to calculate quadratic equations. > This second third still only calculates real roots (leaving aside imaginary > roots). > But thanks to Damien Cassou, and Camille Teruel has a better style. > > A QuadraticEquation is a class used to solve Quadratic Equations of the > form: ax^2 + bx + c = 0, where a is not 0. > > You can set the coefficients using the following method > setQuadraticCoefficient: linearCoefficient: constant:" > > Object subclass: #QuadraticEquation > instanceVariableNames: 'quadraticCoefficient linearCoefficient constant > roots' > classVariableNames: '' > category: 'IS-Math' > > "I create accessors (only getters for each term) and then the following > method to set the terms" > setQuadraticCoefficient: aNumber1 linearCoefficient: aNumber2 constant: > aNumber3 > quadraticCoefficient := aNumber1. > linearCoefficient := aNumber2. > constant := aNumber3 > > > > QuadraticEquation>>calculateRoots > | discriminant | > discriminant := linearCoefficient squared - ( 4 * quadraticCoefficient * > constant). > > discriminant >= 0 > ifFalse: [ ^ self rootsIfNone ] > ifTrue: [ ^ self solveRoots: discriminant] > > > QuadraticEquation>>solveRoots: aTerm > | rootA rootB | > roots := OrderedCollection new. > rootA := (linearCoefficient negated + aTerm sqrt) / (2 * > quadraticCoefficient). > rootB := (linearCoefficient negated - aTerm sqrt) / (2 * > quadraticCoefficient). > roots add: rootA; add: rootB. > ^ roots > > > QuadraticEquation>>rootsIfNone > ^ 'No real solution for the equation' > > > " Example: -3x^2+2x+2=0 > an OrderedCollection(-0.5485837703548636 1.2152504370215302) > > print(it) > > | anEquation | > anEquation := QuadraticEquation new. > anEquation setQuadraticCoefficient: -3 linearCoefficient: 2 constant: 2. > anEquation calculateRoots." > > > > > > > ----- > Nacho > Smalltalker apprentice. > Buenos Aires, Argentina. > -- > View this message in context: http://forum.world.st/Question-on-style-tp4752165p4752369.html > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. >

April 2, 2014

Re: [Pharo-users] Question on style

by Alain Busser

As solving an equation is basically giving the set of its solutions, I prefer to use a set to solve the equation.Instead of the OrederedCollection of course. It simplifies the algorithm if the discriminant is 0, because adding two times the same root yields a set with only one element, not the same element counted twice. The algorithm (much like the one with ordered collection) goes like this: *create an empty set with Set new. *compute the discriminant. *if it is positive, add to the set the "two" roots. *you're done! On Wed, Apr 2, 2014 at 8:45 PM, nacho <0800nacho(a)gmail.com> wrote: > > Hi Nacho, > > I wouldn't throw an error because otherwise one must know beforehand that > an > equation is non-resolvable or must use #on:do: everywhere. > Since that when you ask to resolve an equation you get back a collection of > solutions, a non-resolvable equation could just return an empty collection, > couldn't it?. > If you really want to consider that asking to resolve a non-resolvable > equation is an error, another solution would be to provide a #rootsIfNone: > method. That's a common pattern in Smalltalk. > I would also rename 'checkNegative' to 'discriminant', because it's not > obvious to see what that variable represents at first glance. > > Very good points. I came up with a third draft. This is evolving good! > Thanks > Nacho > > "The idea is to have a class to calculate quadratic equations. > This second third still only calculates real roots (leaving aside imaginary > roots). > But thanks to Damien Cassou, and Camille Teruel has a better style. > > A QuadraticEquation is a class used to solve Quadratic Equations of the > form: ax^2 + bx + c = 0, where a is not 0. > > You can set the coefficients using the following method > setQuadraticCoefficient: linearCoefficient: constant:" > > Object subclass: #QuadraticEquation > instanceVariableNames: 'quadraticCoefficient linearCoefficient > constant > roots' > classVariableNames: '' > category: 'IS-Math' > > "I create accessors (only getters for each term) and then the following > method to set the terms" > setQuadraticCoefficient: aNumber1 linearCoefficient: aNumber2 constant: > aNumber3 > quadraticCoefficient := aNumber1. > linearCoefficient := aNumber2. > constant := aNumber3 > > > > QuadraticEquation>>calculateRoots > | discriminant | > discriminant := linearCoefficient squared - ( 4 * > quadraticCoefficient * > constant). > > discriminant >= 0 > ifFalse: [ ^ self rootsIfNone ] > ifTrue: [ ^ self solveRoots: discriminant] > > > QuadraticEquation>>solveRoots: aTerm > | rootA rootB | > roots := OrderedCollection new. > rootA := (linearCoefficient negated + aTerm sqrt) / (2 * > quadraticCoefficient). > rootB := (linearCoefficient negated - aTerm sqrt) / (2 * > quadraticCoefficient). > roots add: rootA; add: rootB. > ^ roots > > > QuadraticEquation>>rootsIfNone > ^ 'No real solution for the equation' > > > " Example: -3x^2+2x+2=0 > an OrderedCollection(-0.5485837703548636 1.2152504370215302) > > print(it) > > | anEquation | > anEquation := QuadraticEquation new. > anEquation setQuadraticCoefficient: -3 linearCoefficient: 2 > constant: 2. > anEquation calculateRoots." > > > > > > > ----- > Nacho > Smalltalker apprentice. > Buenos Aires, Argentina. > -- > View this message in context: > http://forum.world.st/Question-on-style-tp4752165p4752369.html > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > >

April 2, 2014

Re: [Pharo-users] Question on style

by nacho

Hi Nacho, I wouldn't throw an error because otherwise one must know beforehand that an equation is non-resolvable or must use #on:do: everywhere. Since that when you ask to resolve an equation you get back a collection of solutions, a non-resolvable equation could just return an empty collection, couldn't it?. If you really want to consider that asking to resolve a non-resolvable equation is an error, another solution would be to provide a #rootsIfNone: method. That's a common pattern in Smalltalk. I would also rename 'checkNegative' to 'discriminant', because it's not obvious to see what that variable represents at first glance. Very good points. I came up with a third draft. This is evolving good! Thanks Nacho "The idea is to have a class to calculate quadratic equations. This second third still only calculates real roots (leaving aside imaginary roots). But thanks to Damien Cassou, and Camille Teruel has a better style. A QuadraticEquation is a class used to solve Quadratic Equations of the form: ax^2 + bx + c = 0, where a is not 0. You can set the coefficients using the following method setQuadraticCoefficient: linearCoefficient: constant:" Object subclass: #QuadraticEquation instanceVariableNames: 'quadraticCoefficient linearCoefficient constant roots' classVariableNames: '' category: 'IS-Math' "I create accessors (only getters for each term) and then the following method to set the terms" setQuadraticCoefficient: aNumber1 linearCoefficient: aNumber2 constant: aNumber3 quadraticCoefficient := aNumber1. linearCoefficient := aNumber2. constant := aNumber3 QuadraticEquation>>calculateRoots | discriminant | discriminant := linearCoefficient squared - ( 4 * quadraticCoefficient * constant). discriminant >= 0 ifFalse: [ ^ self rootsIfNone ] ifTrue: [ ^ self solveRoots: discriminant] QuadraticEquation>>solveRoots: aTerm | rootA rootB | roots := OrderedCollection new. rootA := (linearCoefficient negated + aTerm sqrt) / (2 * quadraticCoefficient). rootB := (linearCoefficient negated - aTerm sqrt) / (2 * quadraticCoefficient). roots add: rootA; add: rootB. ^ roots QuadraticEquation>>rootsIfNone ^ 'No real solution for the equation' " Example: -3x^2+2x+2=0 an OrderedCollection(-0.5485837703548636 1.2152504370215302) print(it) | anEquation | anEquation := QuadraticEquation new. anEquation setQuadraticCoefficient: -3 linearCoefficient: 2 constant: 2. anEquation calculateRoots." ----- Nacho Smalltalker apprentice. Buenos Aires, Argentina. -- View this message in context: http://forum.world.st/Question-on-style-tp4752165p4752369.html Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

April 2, 2014

Re: [Pharo-users] Drag and drop items between list views

by Goubier Thierry

Le 02/04/2014 17:40, MartinW a Ã©crit : > Thank you. This is a step forward. The thing is i can't hardcode the > collections. In the end product i don't have two but seven lists, so i must > be able to detect in the acceptDropBlock which is the receiving > collection/list. Ok, the :source is your friend (should be named :target :)); it will give you the Morph (whose model model is the ListModel instance). list1 acceptDropBlock: [ :transfer :event :source :receiver :index | | c1 c2 l1 l2 | l1 := source model model. c1 := l1 listItems. l2 := transfer source model model. c2 := l2 listItems. transfer passenger do: [ :e | c1 insert: e first before: index. c2 remove: e first ]. l1 items: c1. l2 items: c2 ]. This code does not integrate the fact that transfer may be a window (try to drag and drop a window on list1 and you'll see :)) Thierry -- Thierry Goubier CEA list Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s 91191 Gif sur Yvette Cedex France Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95

April 2, 2014

Re: [Pharo-users] Drag and drop items between list views

by Goubier Thierry

Le 02/04/2014 17:40, MartinW a Ã©crit : > Thank you. This is a step forward. The thing is i can't hardcode the > collections. In the end product i don't have two but seven lists, so i must > be able to detect in the acceptDropBlock which is the receiving > collection/list. I'll try to redo a halt in there to see if I can get the target inside the event, and, from the target, be able to get back to the list then :) Thierry > > Goubier Thierry wrote >> Ok, look at that code: >> >> | collection1 collection2 list1 list2 | >> collection1 := #(1 2 3 4 5) asOrderedCollection. >> collection2 := #($a $b $c $d $e) asOrderedCollection. >> list1 := ListModel new. >> list1 items: collection1. >> list1 dragEnabled: true. >> list1 dropEnabled: true. >> list1 acceptDropBlock: >> [ :transfer :event :source :receiver :index | >> transfer passenger do: >> [:e | >> collection1 insert: e first before: index. >> collection2 remove: e first]. >> list1 items: collection1. >> list2 items: collection2 ]. >> list1 openWithSpec. >> list2 := ListModel new. >> list2 items: collection2. >> list2 dragEnabled: true. >> list2 dropEnabled: true. >> list2 acceptDropBlock: [ :transfer :event :source :receiver :index | >> self halt ]. >> ^ list2 openWithSpec >> >> Now you can drag and drop from list2 to list1, and I think the code is >> good for drag and drop of multiple items. > > > > > > -- > View this message in context: http://forum.world.st/Drag-and-drop-items-between-list-views-tp4752285p4752… > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > > > -- Thierry Goubier CEA list Laboratoire des Fondations des SystÃ¨mes Temps RÃ©el EmbarquÃ©s 91191 Gif sur Yvette Cedex France Phone/Fax: +33 (0) 1 69 08 32 92 / 83 95

April 2, 2014