Welcome Guest, you are in: <root>Elements • Login

Page History: Proxy Variables

Compare Page Revisions

28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 Current 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

« Older Revision - Back to Page History - Newer Revision »

Page Revision: 2012/05/09 18:15

Proxy variables are short-cut labels that can be created for selected properties of events and elements. Once a proxy variable is created it becomes accessible as a global variable in all code snippets. This allows to read and modify a status of any EventIDE object at runtime. There is also another class of proxy variables that are defined by user in the Header snippet. Both types of proxy variables are only type of variables that can be used for XAML data binding on the status screen in order to create a dynamic content.

Table of Contents [Hide/Show] Description Classes of Proxy variables     Single Proxy variable     Proxy array variable     Proxy hub variable     User proxy variable Practical Use     Creating a proxy variable     Deleting a proxy variable     Editing a proxy variable     Single Parameters     Parameter Arrays     Multi-Parameter Pool

Description ¶

At the design-time any property of EventIDE events and elements can be modified in the property panel, for example, the black can be chosen as the background color of the currently selected event. The property value, shown in the property panel, is used at the start of an experiment. However, in many scenarios it is also necessary to read and to change object properties at runtime, in user code. For example, you may want to show either green or red text depending on correctness of participant's response. Since an experiment can contain many events and elements a virtual path to modify the Font Color property of the particular Text element may look like this:

[Main_Experiment_Event].[Feedback_Event].[Message_Text_Element].[Font_Color]=new stColor(255,0,0);

Such path is too verbose, difficult to read and and even exhausting to type. EventIDE offers a mechanism of the proxy variables to solve this problem - short-cut labels can be created always for a selected properties of events and elements. In code snippets the proxy variables can be automatically accessed as global variable, whose type matches to the type of underlying property. Then the above code string becomes shorter:

FeedbackMessageColor=new stColor(255,0,0);

or even shorter if you prefer a 'spartan' coding style:

fmc=new stColor(255,0,0);

Note that stColor is a built-in type in EventIDE for color representation.

All proxy variables have to be created manually at design-time, however, it does not take much time because in a typical experiment only a limited number of proxy variables is required. EventIDE provides some browsing and editing capacities for proxy variables in the Proxy Panel.

Proxy variables are allowed not for all properties. For example, the proxies can not be created for all properties that have the 'Design-time Only' usage attribute, like the event title. Also, the proxy variables normally cannot be to created for hardware settings properties, because their modifications at runtime can cause undesirable effects. The other facts about the proxy variables are:

• Only one proxy variable can be created per each property of a particular object.
• Proxy variable always has the same type as its underlying property.
• Proxy variables are automatically removed from an object when its is cut and pasted
  

Classes of Proxy variables ¶

EventIDE allows four different classes of proxy variables.

Single Proxy variable ¶

This type of proxy variables is the most common - a proxy gets linked to a selected property of single EventIDE object.

Proxy array variable ¶

This type of proxy variables is an indexed array that contains several proxy variables, linked to the same property on different objects. The linked objects must be of the same type. The proxy array variable is convenient for enumeration operations in code snippets, e.g. changing the screen positions of multiple targets:

for (int i=0;i<TargetPosition.Length;i++) /// where TargetPosition is an array proxy variable
  {
  TargetPosition[i].X=i*10; 
  TargetPosition[i].Y=368;  
  }

Proxy hub variable ¶

This type of proxy variables allows to modify a property of multiple linked objects at the same time, with a single proxy variable, called hub. The linked objects must be of the same type. The hum proxy variable is convenient when made changes have to be applied to multiple objects, e.g. switching visibility of a group of visual stimuli:

TargetsAreVisible=false; /// TargetsAreVisible is a proxy hub variable linked to multiple objects that change visibility together 

Once the proxy hub variable is created, all underling objects reset the linked properties to the same value automatically and keep them synchronized further, both in run-time and design-time. The hub proxies can not be created for the status (read-only) properties because such properties can not share the same value across different objects.

User proxy variable ¶

Users also can create their own proxy variables that are not linked to EventIDE objects. Any global variable, defined in the Header snippet can be turned to the user proxy variable by adding the 'public' keyword:

// Header code
int A=0; /// this is a usual global variable accessible in all snippets but not in UI.  
public int B=0; /// this is a user proxy variable accessible in all snippets but also in the Proxy Panel. The initial zero value can be overwritten in the Proxy Panel.      

There are two main purposes for creating the user proxy variables

1) The user proxy variable supports change notifications and, therefore, provides self-updating on the Status Screen, when they used in data binding:

   <Textblock Text="{Binding B}"/>
   

This allow allows to monitor the property value in real-time, while an experiment is run. The technique is extremely useful for debugging of snippet code- any variable can be turned into the custom proxy and be monitored on the status screen.

2) the user proxy variables are automatically added into the Proxy Panel, where they can be modified before the start of experiment. The modified values overwrite the initial values defined in the Header snippet code. That can be used to create a run-specific parameter, like participant's name, which can be then set in GUI, before running an experiment.

In the same time, an read-write access to a user proxy variable is slightly slower in code snippets than an access to a normal variable, due to addition of the change notification mechanism. Therefore, creating a large number of the user proxy variables is not recommended.

Practical Use ¶

Creating a proxy variable ¶

To create a single proxy variable, select the object property in the Property Panel by clicking on its name. If property can be linked to a proxy variable, the add button will next to the name:

Creating a proxy variable on the selected property

Alternatively, a proxy variable can be created a command in the main menu. When multiple objects of the same type are selected in the Property Panel, the add proxy button will present two options: create a proxy array or create a proxy hub variable. Once a proxy is created it's shown in the property panel as a label attached to the underlying property.

Deleting a proxy variable ¶

Existing proxy variable can be removed by selecting the underling property in the property panel and pressing the remove button.

Creating a proxy variable on the selected property

Editing a proxy variable ¶

Out Dated text below

Single Parameters ¶

A single parameter is the simplest option and is used to change the parameters of a single element’s property (such as the size or color of an image). Once the parameter has been created and given a unique name, it is necessary to write the appropriate code in the appropriate snippets box, detailing under which conditions, the parameter will take which values. Parameters are often manipulated by predefined random number generators and it is of course necessary to set the limits of this random number generator (see Roulette) as appropriate. Once the possible numbers that could be generated are known, it is necessary to set the values of the parameter relating to each of these outcomes. For example if an experiment uses visual stimuli that could be any one of three colors in any trial, we would have pre-defined our random number generator (using the Main Factors option - see Roulette) to be limited to randomly generating the numbers 1,2 or 3 a fixed amount of times in each block. We would then need to write some code in the snippets, before presentation of the stimulus, stating what the parameter values should be for each of these possible numbers. For example if we have defined the property color of our stimulus element to be the unique parameter "SColor" we would need to write:

if (rNumber==1)
SColor.GDIColor=Color.White;

if (rNumber==2)
SColor.GDIColor=Color.Blue;

if (rNumber==1)
SColor.GDIColor=Color.Black; 

Colors can be defined as either GDI or RGB values – see section (_).

Parameter Arrays ¶

Often it is necessary to change the same property parameters of a number of different elements, at the same time. For example a visual stimulus may consist of three different image elements which can each take on one of three colors. Often in a design of this type it is necessary to have all possible combinations of colors of the three images and a multi-parameter array will allow the easy manipulation of these conditions.

In order to create a multi-parameter array it is necessary to use the pinning function. First navigate to the layer and event window where the first of your elements which you wish to include in the array resides. You can then select the relevant element from the left hand side of the window. Once the element is selected, the property list will be shown on the right hand side of the window and just above this list, a tick box next to the word pinned (see figure 2). Ensure that this tick box is selected, navigate to the next element and repeat this pinning procedure until all the appropriate elements are pinned. Only properties that are common to all the pinned elements can be selected to make a multi-parameter array (See figure 2). Once all the elements are pinned, then the multi-parameter array can be created by either selecting the relevant property to be manipulated and navigating to the Parameters tab and selecting Add as Multi-parameter array or by selecting the small cross that appears when the property is selected (see Figure 2).

Figure 2

The elements that have already been pinned are visible above the properties list. Selecting the property ‘Rotation Angle’ and clicking on the small cross prompts the user to create a multi-parameter array. A previously created multi-parameter array is also visible ‘ColorMask. It can be identified as a multi-parameter array and not merely a single parameter by the 0 after the parameter name.

Once a multiparameter array has been created and named, each of the elements involved will now have the multi-parameter name followed by a unique array number sitting next to the relevant property (see the ColorMask example in Figure 2).

As in the case of the single parameter it is necessary to write the appropriate code in the appropriate snippets box, detailing under which conditions, the parameter will take which values. Parameters are often manipulated by predefined random number generators and it is of course necessary to set the limits of this random number generator (see Roulette Element) as appropriate. Once the possible numbers that could be generated are known, it is necessary to set the values of each of the parameter arrays relating to each of these outcomes. For example if an experiment uses visual stimuli consisting of three elements that could be any one of three colors in any trial, we would have pre-defined our random number generator (using the Main Factors option) to be limited to randomly generating the numbers 1,2 or 3 a fixed amount of times in each block. We would then need to write some code in the snippets, before presentation of the stimulus, stating what the multi-parameter array values should be for each of these possible numbers. For example if we have defined the property ‘color’ of our three stimulus element to be the multi-parameter array ‘SColor’ we would need to write:

if (rNumber==1)
SColor[0].GDIColor=Color.White;
SColor[1] .GDIColor=Color.Blue;
SColor[2].GDIColor=Color.Black;

if (rNumber==2)
SColor[1].GDIColor=Color.White;
SColor[2] .GDIColor=Color.Blue;
SColor[0].GDIColor=Color.Balck;

if (rNumber==3)
SColor[2].GDIColor=Color.White;
SColor[0] .GDIColor=Color.Blue;
SColor[1].GDIColor=Color.Balck; 

This code would ensure that the three element’s properties take on the appropriate values as required given the experimental conditions.

Multi-Parameter Pool ¶

The multi-parameter pool function creates a single point of manipulation for a shared property of different events. If the experimental design requires that the same visual stimulus is used in different events, the color of this stimulus can be manipulated across events with a single line of code.

In order to create a multi-parameter pool it is necessary to use the pinning function. First navigate to the layer and event window where the first of your elements which you wish to include in the array resides. You can then select the relevant element from the left hand side of the window. Once the element is selected, the property list will be shown on the right hand side of the window and just above this list, a tick box next to the word pinned (see figure 3). Ensure that this tick box is selected, navigate to the next element and repeat this pinning procedure until all the appropriate elements are pinned. Only properties that are common to all the pinned elements can be selected to make a multi-parameter pool (See figure 3). Once all the elements are pinned, then the multi-parameter pool can be created by either selecting the relevant property to be pooled and navigating to the Parameters tab and selecting Add as Multi-parameter pool or by selecting the small cross that appears when the property is selected (see Figure 3). It is not possible to create a multi-parameter pool from read-only values (such as statistical properties).

Once the pool has been created, changing the value of this property for one element, will change it for all. A more systematic manipulation can be employed by writing the appropriate code in the appropriate snippets box detailing under which conditions, the parameter pool will take which values. Parameters are often manipulated depending on the number of levels of an experimental factor. These different levels are then generated by predefined random number generators and it is of course necessary to set the limits of this random number generator (see Roulette) as appropriate. Once the possible numbers that could be generated are known, it is necessary to set the values of the parameter pool relating to each of these outcomes. For example if an experiment uses the same visual stimuli across a number of events that could be any one of three colors in any trial, we would have pre-defined our random number generator (using the Main Factors option) to be limited to randomly generating the numbers 1,2 or 3 a fixed amount of times in each block. We would then need to write some code in the snippets, before presentation of the stimulus, stating what the pool values should be for each of these possible numbers. For example if we have defined the property Position of our pooled elements to be the multi-parameter pool ‘SPosition’ we would need to write:

if (rNumber==1)
SPosition =

if (rNumber==2)
SPosition =

if (rNumber==1)
SPosition =

This code will ensure that the stimuli is repositioned as appropriate across all events, without the need to code it for each individual event.