Search Results

Search found 16984 results on 680 pages for 'business logic layer'.

Page 143/680 | < Previous Page | 139 140 141 142 143 144 145 146 147 148 149 150  | Next Page >

  • How to preprocess a Django model field value before return?

    - by Satoru.Logic
    Hi, all. I have a Note model class like this: class Note(models.Model): author = models.ForeignKey(User, related_name='notes') content = NoteContentField(max_length=256) NoteContentField is a custom sub-class of CharField that override the to_python method in purpose of doing some twitter-text-conversion processing. class NoteContentField(models.CharField): __metaclass__ = models.SubfieldBase def to_python(self, value): value = super(NoteContentField, self).to_python(value) from ..utils import linkify return mark_safe(linkify(value)) However, this doesn't work. When I save a Note object like this: note = Note(author=request.use, content=form.cleaned_data['content']) The conversed value is saved into the database, which is not what I wanna see. Would you please tell me what's wrong with this? Thanks in advance.

    Read the article

  • Operation is not valid due to the current state of the object?

    - by Bill
    I am programming in C#; the code was working about a week ago, however it throws an exception and I don't understand at all what could be wrong with it. Var root = new CalculationNode(); -> Throw exception. In the call stack thats the only thing listed, I've been told that it could be that I need a clean build, but I am open to any ideas or suggestions. Thanks, -Bill Update: Exception's Detail System.InvalidOperationException was unhandled by user code Message=Operation is not valid due to the current state of the object. Source=Calculator.Logic StackTrace: at ~.Calculator.Logic.MyBaseExpressionParser.Parse(String expression) in ~\Source\Calculator.Logic\MyBaseExpressionParser.cs:line 44 at ~.Calculator.Logic.Tests.MyBaseCalculatorServiceTests.BasicMathDivision() in ~\Projects\Tests\Calculator.Logic.Tests\MyBaseCalculatorServiceTests.cs:line 60 InnerException: CalculationNode's code: public sealed calss CalculationNode { public CalculationNode() { this.Left = null; this.Right = null; this.Element = new CalculationElement(); } public CalculationNode Left {get;set;} public CalculationNode Right {get;set;} public CalculationElement Element {get; set;} } CalculationElement's code: public sealed class CalculationElement { public CalculationElement() { Value = string.Empty; IsOperator = false; } public string Value {get; set} public bool IsOperator {get; set} }

    Read the article

  • 3 Dimensional Array

    - by George Johnston
    I don't know why I'm experiencing so much trouble with this, but I would like to have an array that basically represents a layer number and x,y coordinates so I could essentially say, int i = array[layer,x,y] and get the corrisponding value per layer. I create the array.. int[,,] myarray ...initialize it myarray = new int[0,width, height]; ...and it blows up when try and grab a value. int n = myarray[0,1,1] What am I missing?

    Read the article

  • Display an Image using C# in Web Application

    - by Josh
    Hello again, I have a Web Application that I built a C# class in that generates a Report. This report takes nearly 40 seconds to generate because it searches hundreds of folders for certain files. So I was hoping there was a way to display a "Loading.." icon as this report is generating. I have a gif stored in my Images folder that would be perfect. The research that I've done at this point mostly talks about picturebox's and image controls that can hold the image but I was hoping there was just a way of displaying the image above the report while its being created. The Web Application is from a Web ADF Geocortex website and again I created a C# class that generates this report. Below is some code that may help. /// <summary> /// Generates HTML for the current report using the data in /// the given table. /// </summary> /// <param name="reportLayers">The layers to include in the report.</param> /// <returns> public override string GenerateReportHtml(List<ReportLayer> reportLayers) { StringBuilder htmlString = new StringBuilder(); StringWriter stringWriter = new StringWriter(htmlString); HtmlTextWriter writer = new HtmlTextWriter(stringWriter); string holdAPI = ""; List<string> exclusions = GetExcludedFields(); foreach (ReportLayer layer in reportLayers) { string[] strFiles = null; Boolean val = false; if (layer.Layer.Name == "Bottom Hole Location (OP)") writer.RenderBeginTag(HtmlTextWriterTag.P); // <p> writer.RenderBeginTag(HtmlTextWriterTag.Strong); // <strong> writer.Write(layer.Layer.Name); writer.RenderEndTag(); // end </strong> writer.RenderEndTag(); // end </p> writer.WriteBreak(); // <br /> foreach (ReportFeature feature in layer.ReportFeatures) { // each feature will be in a table writer.RenderBeginTag(HtmlTextWriterTag.Table); // <table> foreach (ReportField field in feature.ReportFields) { string fieldName = field.Alias; if (!exclusions.Contains(fieldName)) {

    Read the article

  • What's the workflow of Continuous Integration With Hudson?

    - by Satoru.Logic
    Hi, all. I am referred to Hudson today. I have heard about continuous integration before, but I have no idea what the heck is a ci-server. Hudson is really easy to install in Ubuntu and in several minutes I managed to set up an instance of it. But I don't quite understand the workflow of a ci-server, or how am I supposed to use it? Please tell me if you have experience about ci, thanks in advance.

    Read the article

  • WCF Architectural Debate

    - by Matt Ruwe
    When building a WCF service for a large scale application, which is better: In both cases, assume that the business logic layer is separated into a different assembly. Using your business logic layer as a service implementation i.e. with no code behind and no wrapper <%@ ServiceHost Language="CS" Service="MyApp.BusinessLogic.BusLogicImpl" %> or Using the codebehind of the WCF service that wraps calls into your business logic layer. <%@ ServiceHost Language="CS" Service="MyApp.WebServiceHost.Service" CodeBehind="Service.svc.cs" %>

    Read the article

  • Where about should my main class be created in a project?

    - by Dan
    The problem is where a class should be created in my code. An example is I have a UI class and a main logic class that controls other objects. Should the main logic class create the UI object, or should the UI object create the instance of the main logic class? An explanation of which method is best and why would be ideal. Thanks.

    Read the article

  • Combination of JFreeChart with JXLayer with JHotDraw

    - by Yan Cheng CHEOK
    Recently, I had use JXLayer, to overlay two moving yellow message boxes, on the top of JFreeChart http://yccheok.blogspot.com/2010/02/investment-flow-chart.html I was wondering, had anyone experience using JXLayer + JHotdraw, to overlay all sorts of figures (Re-sizable text box, straight line, circle...), on the top of JFreeChart. I just would like to add drawing capability, without changing the JFreeChart source code. So that, JStock's user may draw trending lines, annotation text on their favorite stock charting. The code skeleton is as follow : // this.chartPanel is JFreeChartPanel final org.jdesktop.jxlayer.JXLayer<ChartPanel> layer = new org.jdesktop.jxlayer.JXLayer<ChartPanel>(this.chartPanel); this.myUI = new MyUI<ChartPanel>(this); layer.setUI(this.myUI); public class MyUI<V extends javax.swing.JComponent> extends AbstractLayerUI<V> { @Override protected void paintLayer(Graphics2D g2, JXLayer<? extends V> layer) { // Previous, I am using my own hand-coded, to draw the yellow box // // Now, How can I make use of JHotDraw at here, to draw various type of // figures? } @Override protected void processMouseEvent(MouseEvent e, JXLayer<? extends V> layer) { // How can I make use of JHotDraw at here? } @Override protected void processMouseMotionEvent(MouseEvent e, JXLayer<? extends V> layer) { // How can I make use of JHotDraw at here? } } As you see, I already got Graphics2D g2 from paintLayer method. How is it possible that I can pass the Graphics2D object to JHotDraw, and let JHotDraw handles all the drawing. My experience in using JHotDraw are with org.jhotdraw.draw.DefaultDrawingView org.jhotdraw.draw.DefaultDrawingEditor I am able to use them to draw various figures, by clicking on the toolbar and click on drawing area. How is it possible I can use DefaultDrawingView and DefaultDrawingEditor within MyUI's paintLayer? Also, shall I let MyUI handles the mouse event, or JHotDraw? Sorry, I start getting confused.

    Read the article

  • Difference in linq-to-sql query performance using GenericRespositry

    - by Neil
    Given i have a class like so in my Data Layer public class GenericRepository<TEntity> where TEntity : class { [System.ComponentModel.DataObjectMethod(System.ComponentModel.DataObjectMethodType.Select)] public IQueryable<TEntity> SelectAll() { return DataContext.GetTable<TEntity>(); } } I would be able to query a table in my database like so from a higher layer using (GenericRepositry<MyTable> mytable = new GenericRepositry<MyTable>()) { var myresult = from m in mytable.SelectAll() where m.IsActive select m; } is this considerably slower than using the usual code in my Data Layer using (MyDataContext ctx = new MyDataContext()) { var myresult = from m in ctx.MyTable where m.IsActive select m; } Eliminating the need to write simple single table selects in the Data layer saves a lot of time, but will i regret it?

    Read the article

  • How to perform duplicate key validation using entlib (or DataAnnotations), MVC, and Repository pattern

    - by olivehour
    I have a set of ASP.NET 4 projects that culminate in an MVC (3 RC2) app. The solution uses Unity and EntLib Validation for cross-cutting dependency injection and validation. Both are working great for injecting repository and service layer implementations. However, I can't figure out how to do duplicate key validation. For example, when a user registers, we want to make sure they don't pick a UserID that someone else is already using. For this type of validation, the validating object must have a repository reference... or some other way to get an IQueryable / IEnumerable reference to check against other rows already in the DB. What I have is a UserMetadata class that has all of the property setters and getters for a user, along with all of the appropriate DataAnnotations and EntLib Validation attributes. There is also a UserEntity class implemented using EF4 POCO Entity Generator templates. The UserEntity depends on UserMetadata, because it has a MetadataTypeAttribute. I also have a UserViewModel class that has the same exact MetadataType attribute. This way, I can apply the same validation rules, via attributes, to both the entity and viewmodel. There are no concrete references to the Repository classes whatsoever. All repositories are injected using Unity. There is also a service layer that gets dependency injection. In the MVC project, service layer implementation classes are injected into the Controller classes (the controller classes only contain service layer interface references). Unity then injects the Repository implementations into the service layer classes (service classes also only contain interface references). I've experimented with the DataAnnotations CustomValidationAttribute in the metadata class. The problem with this is the validation method must be static, and the method cannot instantiate a repository implementation directly. My repository interface is IRepository, and I have only one single repository implementation class defined as EntityRepository for all domain objects. To instantiate a repository explicitly I would need to say new EntityRepository(), which would result in a circular dependency graph: UserMetadata [depends on] DuplicateUserIDValidator [depends on] UserEntity [depends on] UserMetadata. I've also tried creating a custom EntLib Validator along with a custom validation attribute. Here I don't have the same problem with a static method. I think I could get this to work if I could just figure out how to make Unity inject my EntityRepository into the validator class... which I can't. Right now, all of the validation code is in my Metadata class library, since that's where the custom validation attribute would go. Any ideas on how to perform validations that need to check against the current repository state? Can Unity be used to inject a dependency into a lower-layer class library?

    Read the article

  • How to secure authiorization of methods

    - by Kurresmack
    I am building a web site in C# using MVC.Net How can I secure that no unauthorized persons can access my methods? What I mean is that I want to make sure that only admins can create articles on my page. If I put this logic in the method actually adding this to the database, wouldn't I have business logic in my data layer? Is it a good practise to have a seperate security layer that is always in between of the data layer and the business layer to make? The problem is that if I protect at a higher level I will have to have checks on many places and it is more likely that I miss one place and users can bypass security. Thanks!

    Read the article

  • How to efficiently handle Where and OrderBy clauses

    - by Goran
    My business layer passes all the required information to UI layer. From what I have read, in general, best practice is to send fetched data to UI layer, and to avoid passing queries like ObjectQuery. My problem with this approach is next: If I am to make a flexible business layer, then I should allow UI to sort the data anyway it requires. Fetching sorted data from database, and then resorting them in UI is kind of bad practice for me, so the only way is to somehow So what are my options? Is there a way to make it like this: public void OrderByMethod(params ...) { .... } so I can call it like this: OrderByMethod(MyEntity.Property1, MyEntity.Property2 descending....); Thanks, Goran

    Read the article

  • Get wrong PATH_INFO after rewriting in lighttpd

    - by Satoru.Logic
    In my lighttpd config file, I have a rewrite rule like this: $HTTP["host"] == "sub.example.com" { url.rewrite = ( "^/(.*)" => "/sub/$1" ) } So when a user visits http://sub.example.com, she's actually visiting http://example.com/sub. The problem is that the PATH_INFO seems wrong, URL: http://sub.example.com/extra PATH_INFO: expected: /extra what I get: /sub/extra Now whenever I call request.get_path(), it returns something like http://sub.example.com/sub/extra, which is not what I want. Of course, I can just override the get_path method of the request class, but I wonder if there is a simpler way like changing the lighttpd config?

    Read the article

  • Designing a class in such a way that it doesn't become a "God object"

    - by devoured elysium
    I'm designing an application that will allow me to draw some functions on a graphic. Each function will be drawn from a set of points that I will pass to this graphic class. There are different kinds of points, all inheriting from a MyPoint class. For some kind of points it will be just printing them on the screen as they are, others can be ignored, others added, so there is some kind of logic associated to them that can get complex. How to actually draw the graphic is not the main issue here. What bothers me is how to make the code logic such that this GraphicMaker class doesn't become the so called God-Object. It would be easy to make something like this: class GraphicMaker { ArrayList<Point> points = new ArrayList<Point>(); public void AddPoint(Point point) { points.add(point); } public void DoDrawing() { foreach (Point point in points) { if (point is PointA) { //some logic here else if (point is PointXYZ) { //...etc } } } } How would you do something like this? I have a feeling the correct way would be to put the drawing logic on each Point object (so each child class from Point would know how to draw itself) but two problems arise: There will be kinds of points that need to know all the other points that exist in the GraphicObject class to know how to draw themselves. I can make a lot of the methods/properties from the Graphic class public, so that all the points have a reference to the Graphic class and can make all their logic as they want, but isn't that a big price to pay for not wanting to have a God class?

    Read the article

  • MVC2 Data Annotation Buddy Classes Doesn't seem to work when Classes and EOM Model is in separate Project

    - by Danish Ali
    Dear All Iam new to MVC2 and having a little problem with implementing validation via buddy classes. Iam using Repository pattern with dependency injection. My Entity Object Model is in Data Layer Project and Buddy Classes are in Business Layer project and MVC 2 Project as a separate Presentation Layer Project. Can any one help me out with implementing Buddy Classes in this Architecture. Thanks and Regards Dani

    Read the article

  • How do I stack images to simulate depth using Core Animation?

    - by Jeffrey Berthiaume
    I have a series of UIImages with which I need to simulate depth. I can't use scaling because I need to be able to rotate the parent view, and the images should look like they're stacked visibly in front of each other, not on the same plane. I made a new ViewController-based project and put this in the viewDidLoad (as well as attached three 120x120 pixel images named 1.png, 2.png, and 3.png): - (void)viewDidLoad { // display image 3 UIImageView *three = [[UIImageView alloc] initWithImage:[UIImage imageNamed:@"3.png"]]; three.center = CGPointMake(160 + 60, 240 - 60); [self.view addSubview:three]; // rotate image 3 around the z axis // THIS IS INCORRECT CATransform3D theTransform = three.layer.transform; theTransform.m34 = 1.0 / -1000; three.layer.transform = theTransform; // display image 2 UIImageView *two = [[UIImageView alloc] initWithImage:[UIImage imageNamed:@"2.png"]]; two.center = CGPointMake(160, 240); [self.view addSubview:two]; // display image 1 UIImageView *one = [[UIImageView alloc] initWithImage:[UIImage imageNamed:@"1.png"]]; one.center = CGPointMake(160 - 60, 240 + 60); [self.view addSubview:one]; // rotate image 3 around the z axis // THIS IS INCORRECT theTransform = one.layer.transform; theTransform.m34 = 1.0 / 1000; one.layer.transform = theTransform; // release the images [one release]; [two release]; [three release]; // rotate the parent view around the y axis theTransform = self.view.layer.transform; theTransform.m14 = 1.0 / -500; self.view.layer.transform = theTransform; [super viewDidLoad]; } I have very specific reasons why I'm not using an EAGLView and why I'm not loading the images as CALayers (i.e. why I'm using UIImageViews for each one). This is just a quick demo that I can use to work out exactly what I need in my parent application. Is there some matrix way to translate these 2d images along the z-axis so they will look like what I'm trying to represent? I've gone through the other StackOverflow articles as well as the Wikipedia references, and have not found what I'm looking for -- although I might not necessarily be using the right terms for what I'm trying to do.

    Read the article

  • PHP Frontpage/Page controller

    - by atno
    I using the following as Frontpage/Page Controller(s) and it's working ok so far, except two problems I'm facing which as you can see are the $pages array and the switch, which are actually much much longer as the one I've pasted here. Everytime there is a need for a new page controller I have to add it to $pages array and to switch which makes that list very long. How would you overcome this problem and do you see any other improvement on this code? loadLogic() in page controllers it is used to get functions under pages/controllername/logic/function.php. Frontpage Controller - index.php: include 'common/common.php'; if(!isset($_GET['p']) OR $_GET['p'] == ''){ $_GET['p'] = 'home'; header('Location: index.php?p=home'); } $pages = array('home','register','login','logout','page1','page2','page3'); $_GET['p'] = trim($_GET['p']); if(isset($_GET['p'])){ if(in_array($_GET['p'], $pages)){ switch ($_GET['p']) { case 'home': include 'home.php'; break; case 'register': include 'register.php'; break; case 'login': include 'login.php'; break; case 'logout': include 'logout.php'; break; case 'page1': include 'page1.php'; break; case 'page2': include 'page2.php'; break; case 'page3': include 'page3.php'; break; } }else{ echo '404!'; } } Page Controller - {home,register,login,logout,page1,page2,page3}.php: include 'tpl/common/header.php'; contentStart(); if(isset($_SESSION['logged'])){ loadLogic('dashboard'); }else{ loadLogic('nologin'); } //Display login form in logic page instead links // if(!isset($_SESSION['logged'])){ contentEnd(); loadLogic('nologinForm'); }else{ contentEnd(); include'tpl/common/rcol.php'; } include 'tpl/common/footer.php'; function loadLogic(): function loadLogic($logic) { $path = dirname(__DIR__) . '/pages'; $controller = preg_split('/&/',$_SERVER['QUERY_STRING']); $controller = trim($controller[0],"p="); $logicPath = 'logic'; $logic = $logic . '.php'; $err = 0; $logicFullPath = $path.'/'.$controller.'/'.$logicPath.'/'.$logic; if($err == '0'){ include "$logicFullPath"; } } Folder Structure: projectName | ---> common | ---> pages | | | --->home | | | --->register | | | --->login | | | --->logout | | | --->page1 | | | --->page2 | | | --->page3 | ---> tpl | | | ---> common | --> home.php | --> register.php | --> login.php | --> logout.php | --> page1.php | --> page2.php | --> page3.php

    Read the article

  • what should I take into consideration when choosing a session implementation?

    - by Satoru.Logic
    Hi, all. Yesterday, my supervisor told me that tmp-file-based session should be THE answer to session implementation, and I should abandon any idea of making sessions persistent. He argues that file-based session is much faster and eaiser to use than other choices like db-based or memcached-based implementations. What he said was really a shock to my past learning experience, so please tell me whether he was wrong and why? Thanks in advance.

    Read the article

  • Flash error 1084: "Syntax error"

    - by Elliot Broomhall
    Hi I'm getting two error messages in Flash when using actionscropt 3.0 "Topbar,Layer 'Action Layer',Frame 1,line 12 1084: syntax error: expection semicolon before add. "Topbar,Layer 'Action Layer',Frame 1,line 12 1084: syntax error: expection rightbrace before semicolon Here is my code could anyone give some insight to what is actually happening thanks and help on rectifying the issue thanks. clip = Number(random(7)) + 1; while (Number(clip) <= 7) { clip = Number(clip) + 1; Scale = Number(random(80)) + 1; setProperty("/star", _x, Number(random(800)) + 10); setProperty("/star", _rotation, Number(random(330)) + 50); setProperty("/star", _xscale, Scale); setProperty("/star", _yscale, Scale); setProperty("/star", _y, Number(random(800)) + 50); n = Number(n) + 1; bn = "star" add n; duplicateMovieClip("star", bn, n); set(bn add ":n", n); } // end while clip = "0";

    Read the article

  • Eclipse Plugin: Enablement of an Action based on the current selection

    - by Itay
    I am using the org.eclipse.ui.popupMenus extension point for adding a sub-menu whose Action that is bounded to the following class: public class MyAction implements IObjectActionDelegate { private Logic logic = Logic.getInstance(); // Singleton public void setActivePart(IAction a, IWorkbenchPart targetPart) { // Nothing here } public void run(IAction a) { // Do something... } public void selectionChanged(IAction a, ISelection s) { a.setEnabled(logic.isEnabled(s)); } } This action is working correctly in most cases (including the call a.setEnabled() in selectionChanged()). My problem at the very first time my action is being invoked. The selectionChanged method is called only after the menu item has been displayed (and not when the user has made the selection) which means that the call to a.setEnabled() will have no affect. Any ideas on how to make my action receive selectionChanged() notifications even before the fist time it is being invoked?

    Read the article

  • Combining two .png images into one image using .NET

    - by Omega
    I have two (actually many) .png images in my application. Both have transparent areas here and there. I want, in my application, to take both images, combine them, and display the result in a picture box. Later I want to save the result through a button. So far I managed to find the two images and combine them, but it seems the transparency thing won't work. I mean, if you put one image over another, only the top image is visible as the result because, apparently, the image's background is a plain white box. Which is not. Here is a bit of my code: Dim Result As New Bitmap(96, 128) Dim g As Graphics = Graphics.FromImage(Result) Dim Name As String For Each Name In BasesCheckList.CheckedItems Dim Layer As New Bitmap(resourcesPath & "Bases\" & Name) For x = 0 To Layer.Width - 1 For y = 0 To Layer.Height - 1 Result.SetPixel(x, y, Layer.GetPixel(x, y)) Next Next Layer = Nothing Next resourcesPath is the path to my resources folder. Bases is a folder in it. And Name is the image's name. Thank you.

    Read the article

  • UIView animation problem

    - by FoJjen
    I try to make a UIImageView open like a hatch, and under the "hatchView" its anoter imageView but when I start the animation its like the view under gets over and cover the animation, When the animation gets over the imageView the its getting visibel. its works fine if a remove the imageView thats under. Here is the code for animation view. - (void) touchesEnded:(NSSet *)touches withEvent:(UIEvent *) event { if (!isHatchOpen) { if (hatchView.layer.anchorPoint.x != 0.0f) { hatchView.layer.anchorPoint = CGPointMake(0.0f, 0.5f); hatchView.center = CGPointMake(hatchView.center.x - hatchView.bounds.size.width/2.0f, hatchView.center.y); } UITouch *touch = [touches anyObject]; CGPoint location = [touch locationInView:hatchView]; [UIView beginAnimations:nil context:NULL]; [UIView setAnimationDuration:1]; [UIView setAnimationCurve:UIViewAnimationCurveEaseInOut]; [UIView setAnimationBeginsFromCurrentState:YES]; float pct = location.x / 320.0f; float rad = acosf(pct); CATransform3D transform = CATransform3DMakeRotation(-rad+5, 0.0f, -1.0f, 0.0f); transform.m14 = (1.0 / -2000) * acosf(pct); hatchView.layer.transform = transform; [UIView commitAnimations]; isHatchOpen = YES; } else { if (hatchView.layer.anchorPoint.x != 0.0f) { hatchView.layer.anchorPoint = CGPointMake(0.0f, 0.5f); hatchView.center = CGPointMake(hatchView.center.x - hatchView.bounds.size.width/2.0f, hatchView.center.y); } [UIView beginAnimations:nil context:NULL]; [UIView setAnimationDuration:1]; [UIView setAnimationCurve:UIViewAnimationCurveLinear]; [UIView setAnimationBeginsFromCurrentState:YES]; CATransform3D transform = CATransform3DMakeRotation(0, 0.0f, -1.0f, 0.0f); transform.m14 = 0; hatchView.layer.transform = transform; // Commit the changes [UIView commitAnimations]; isHatchOpen = NO; } } And here a add the Views hatchView = [[UIImageView alloc] initWithFrame:CGRectMake(0, 1, frame.size.width-1, frame.size.height-2)]; [self addSubview:hatchView]; imageView = [[UIImageView alloc] initWithFrame:CGRectMake(1, 1, frame.size.width-2, frame.size.height-2)]; imageView.contentMode = UIViewContentModeScaleAspectFit; [self insertSubview:imageView belowSubview:hatchView];

    Read the article

  • Questions about the Backpropogation Algorithm

    - by Colemangrill
    I have a few questions concerning backpropogation. I'm trying to learn the fundamentals behind neural network theory and wanted to start small, building a simple XOR classifier. I've read a lot of articles and skimmed multiple textbooks - but I can't seem to teach this thing the pattern for XOR. Firstly, I am unclear about the learning model for backpropogation. Here is some pseudo-code to represent how I am trying to train the network. [Lets assume my network is setup properly (ie: multiple inputs connect to a hidden layer connect to an output layer and all wired up properly)]. SET guess = getNetworkOutput() // Note this is using a sigmoid activation function. SET error = desiredOutput - guess SET delta = learningConstant * error * sigmoidDerivative(guess) For Each Node in inputNodes For Each Weight in inputNodes[n] inputNodes[n].weight[j] += delta; // At this point, I am assuming the first layer has been trained. // Then I recurse a similar function over the hidden layer and output layer. // The prime difference being that it further divi's up the adjustment delta. I realize this is probably not enough to go off of, and I will gladly expound on any part of my implementation. Using the above algorithm, my neural network does get trained, kind of. But not properly. The output is always XOR 1 1 [smallest number] XOR 0 0 [largest number] XOR 1 0 [medium number] XOR 0 1 [medium number] I can never train the [1,1] [0,0] to be the same value. If you have any suggestions, additional resources, articles, blogs, etc for me to look at I am very interested in learning more about this topic. Thank you for your assistance, I appreciate it greatly!

    Read the article

  • Basic data alignment question

    - by Broken Logic
    I've been playing around to see how my computer works under the hood. What I'm interested in is seeing is what happens on the stack inside a function. To do this I've written the following toy program: #include <stdio.h> void __cdecl Test1(char a, unsigned long long b, char c) { char c1; unsigned long long b1; char a1; c1 = 'b'; b1 = 4; a1 = 'r'; printf("%d %d - %d - %d %d Total: %d\n", (long)&b1 - (long)&a1, (long)&c1 - (long)&b1, (long)&a - (long)&c1, (long)&b - (long)&a, (long)&c - (long)&b, (long)&c - (long)&a1 ); }; struct TestStruct { char a; unsigned long long b; char c; }; void __cdecl Test2(char a, unsigned long long b, char c) { TestStruct locals; locals.a = 'b'; locals.b = 4; locals.c = 'r'; printf("%d %d - %d - %d %d Total: %d\n", (long)&locals.b - (long)&locals.a, (long)&locals.c - (long)&locals.b, (long)&a - (long)&locals.c, (long)&b - (long)&a, (long)&c - (long)&b, (long)&c - (long)&locals.a ); }; int main() { Test1('f', 0, 'o'); Test2('f', 0, 'o'); return 0; } And this spits out the following: 9 19 - 13 - 4 8 Total: 53 8 8 - 24 - 4 8 Total: 52 The function args are well behaved but as the calling convention is specified, I'd expect this. But the local variables are a bit wonky. My question is, why wouldn't these be the same? The second call seems to produce a more compact and better aligned stack. Looking at the ASM is unenlightening (at least to me), as the variable addresses are still aliased there. So I guess this is really a question about the assembler itself allocates the stack to local variables. I realise that any specific answer is likely to be platform specific. I'm more interested in a general explanation unless this quirk really is platform specific. For the record though, I'm compiling with VS2010 on a 64bit Intel machine.

    Read the article

< Previous Page | 139 140 141 142 143 144 145 146 147 148 149 150  | Next Page >