Private data members vs. local variables

Hi
This is more of a basic design question.
I have a code that retrieves a lot of data from a text file and puts them into the database. I have one function per table for inserting rows which are called from various functions in the code. I have all the functions (calling and called) in the same java class.
The number of variables that are needed to hold the data are approx 40. I was thinking if I should have them as class variables and update them and just access the values in the functions where I do the database updates or should I have them as local variables and pass them when a function is being called. Or should I be putting them into some data structure and pass the data structure.
I want to know which method would be memory efficient and would not hurt the performance of my code.
Please do let me know your valuable opinions.
I would greatly appreciate any thoughts on this.
Thanks

I would strongly advice against using fields as a way to get around long method invocations.
For one thing, it's not using fields as object state.
Second, it's a really good source for bugs. I've seen it happen. The "bucket" fields (to coin a phrase; I mean fields used to bypass long argument lists) become messy, unreliable data, and next thing you know some method is breaking other methods. This becomes especially dangerous when there's multithreading, but you don't need multithreading to get bugs from this.
If you have a lot of values you need to pass around, then two possibilities occur to me. One is that these values are configuration data, in which case they probably do qualify as object state and maybe you can use fields. The other possibility is that you need to refactor your code; all those values may need to live in one or more new classes that you need to write. Another refactoring may be the order in which you perform the steps in your algorithm.

Similar Messages

What is it about data types of local variables?

Hey there!
I'm trying hard to access the booleans, whose refnums got stored into an array. As the example show, everything's fine unless I start using local variables: I can access the data in the topmost FOR loop. As soon as variables come into play, it seems as if I have to explicitly typecast the data. Is it just me? There has to be a less complicated way... can anybody give me a hint?
Kind regards,
Severin
Attachments:
data types of local variables.jpg ‏96 KB
data types of local variables_7.11.vi ‏39 KB

Hi Severin,
The output from that property node is of variant data type. You can convert it to boolean by using the Variant to Data VI. See attached example
Sarah
Applications Engineer | National Instruments | UK & Ireland
Attachments:
data types of local variables_7.11.vi ‏41 KB

Data Acquisition - using local variables to write data to a file

Hello,
I am running a Data Acquisition vi (currently in LabVIEW 7.1 but soon to be updated to 8.2) that collects ~100 parameters of data from several sources contained in a while loop. The current configuration (which I did not write) uses very few subVIs and writes to ~100 local variables to store each parameter. It then reads all the local variables and builds an array of all the strings, converts then to a spreadsheet string, then uses the write characters to file function to append to a datafile. I am trying to clean things up and have came up with subVIs to collect the data from the following sources:
8 serial port sources collecting btwn 8 and 20 parameters each
~15 thermocouple readings
~10 analog inputs
~20 parameters read off an ARINC 429 bus.
I have come up with a subVI to read each of the sources and have placed the subVIs in the while loop. Each subVI outputs the data that it collects in array or cluster form. I was wondering how best to write each parameter to a CSV file at between 1 and 10 Hz. Should I write each subVI output to a LV and then read them off as was done before (the difference being that I have reduced the # of LVs to ~10 vs >100?
I should add that precise timing is not that important, so if all the subVIs are not collecting simultaneously (which I understand that they won't be), it does not really matter.
Thanks.

Hi jilla,
jilla wrote:
What I think that you are saying is to turn the outputs of the 4 subVIs into inputs of a 5th subVI that writes to the data file. Correct?
Yes. It may sound like a fine-point, but I beileve it's better to create a VI specifically for formatting data - in your example, 4 arrays IN, a single string OUT. Then write the string to file as a seperate operation. GUI-displayed data can go through a similar transformation, the four arrays wired to a subVI which builds output-structures specifically for display. It's a beginner's mistake to put lots of individual controls and indicators on the screen when groups of them are naturally related (in an object-oriented sense.) Use clusters to group related controls - this will keep the diagram much cleaner.
One more question: at what point (either # of data points or frequency of data collection) does it become necessary to use queues? Thanks.
Well, there's not really a clearly definable "point". I'd say if your update-rate climbs above 100Hz, or you witness poor program or system performance, then it's time. The scenario you've described is a fairly simple acquire/display&log loop - and simple is good. Then-again people can't see/react-to updates faster than about 10Hz - so it doesn't make sense to sacrifice performance - if performance becomes an issue.
Re: queues: Queues are sometimes used to buffer data that's "produced" in one place and "consumed" in another.
Here, if/when logging data, you're logging with every DAQ. I wouldn't recommend using a queue to transport data from a "DAQ loop" to a "Logging-loop" - those functions can be in the same loop. Should/could a queue be used to get data from a "DAQ loop" to update the GUI at a lower frequency? Sure, but a Notifier might be a better choice. Further, in the (simple?) program you've described, you might use a case structure (True/False) to only update FP indicators every "X" iterations - a simple solution that doesn't require Queues or Notifiers.
Cheers!
"Inside every large program is a small program struggling to get out." (attributed to Tony Hoare)

How to define a TYPE REF TO data into my local variable/structure?

Hello,
I am trying to implementing a BADI, in its signature part,
c_accit TYPE REF TO data.
My pseudo code goes like,
DATA: l_itm_details TYPE REF TO data.
l_itm_details = c_accit.
l_itm_details-koart = 'D' =====> here am getting error!
* Do processing
ENDIF.
Here am getting error that, l_itm_details is not a structure! Pls. let me know how to fix it?
Thank you

Hello,
c_accit TYPE REF TO data.
C_ACCIT is a "data reference" parameter. In order to access its components you have to "de-reference" it!
FIELD-SYMBOLS: <l_itm_details> TYPE ANY,
<l_value> TYPE ANY.
ASSIGN c_accit->* TO <l_itm_details>."De-reference the data reference
ASSIGN COMPONENT 'KOART' OF STRUCTURE <l_itm_details> TO <l_value>.
If you know the structure of the data reference variable you can define your field-symbol <l_itm_details> of that type directly, else you can define a generic type as mentioned in the code snippet.
BR,
Suhas

How can I Change data in a type def control containing a Xcontrol with a local variable

Hello
I made a Xcontrol and I inserted this control in a type def.
When I want to change the control's data with a local variable in a VI, the VI change nothing. The change of data isn't perform
How can I correct this issue?
Thanks for your help
Solved!
Go to Solution.

Hello,
What's your LabVIEW version ? Do you have a simple example program which demonstrates this behavior ?
I found another discussions related to your issues with Xcontrols:
updating type defs in Xcontrol Facade
No Data Change event generated for a XControl in a Type Def
XControl facede.vi
Hope this helps.
Regards,
Steve M.
National Instruments France
#adMrkt{text-align: center;font-size:11px; font-weight: bold;} #adMrkt a {text-decoration: none;} #adMrkt a:hover{font-size: 9px;} #adMrkt a span{display: none;} #adMrkt a:hover span{display: block;}
>> Vidéo-t'chats de l'été : présentations techniques et ingénieurs pour répondre à vos questions

Best way to remove local variables

Hello,
I have been told that I need to remove all the local variables in my VI, because they are slow. I can not seem to find a good option for replacing them but keeping the same functionality.
What I am using the local variables for is to update 3 different graphs on the front panel from 3 different sources. Each of these sources are located in while loops so that data can be acquired asynchronously. In closing these 3 while loops is a case statement. The case statement is for switching between acquiring and saving the data or just acquiring data. The local variables are used to update the indicators and graphs from the acquiring and saving data case. It is these local variables that I need to remove.
Does anyone have any suggestions?
-- Z

Zurvan wrote:
I have been told that I need to remove all the local variables in my VI, because they are slow.
Who told you that?
As others have said, local variables typically don't cause slowdown (They cause additional datacopies in memory, have the potential of causing race conditions and cause messy code).
What is "slow" in your definition? (missing data, cannot keep up with the instrument, slugging UI, etc.).
What is the CPU use when your program runs?
So apparently you have a loop that updates all indicators from local variables. What is the loop rate of that display loop? Do you spin that millions of times per second or pace it at a reasonable rate? How much data is in the graphs?
Remember, a graph does not need to update unless the data changes.
LabVIEW Champion . Do more with less code and in less time .

Local variable fail when using SIT

Dear all,
I am using SIT (Simulink Interface Toolkit) to obtain data generated by the model of Simulink. I need to use the local variable of the indicator to get the data out.
The interface with the mappings seems to work fine. The problemn is that the local variable of the waveform （this waveform is an indicater which mapped to the output signal of simulink and it can displays the signal from simulink）didn't work. I cann't obtain any data from the local variable, but the waveform which the local variable belonges to seems to work fine.
I also found an interesting thing that, when I hightlighting the execution process, the local variable seems like didn't work at all, because there is no data flows from it.
The software version is Matlab simulink R2011b, Labview 2011.
Below is a simple test model in simulink. It is a sine signal.
Below is the local variable from the waveform which mapped to the signal of simulink. To check there is any data in the local variable, I put a waveform to check.

Dataflow.
The local variable of the waveform is going to be empty to start. The writing of the local variable called Waveform to the indicator called Local Variable is going to happen right away. Only later does the waveform get any data. But by then the first part has already exectued never to be run again.
I would recommend looking at the online LabVIEW tutorials
LabVIEW Introduction Course - Three Hours
LabVIEW Introduction Course - Six Hours

Local variable in Data Form

Hi All,
I have a dataform in which Entity Dim is in Page. In a Business Rule attached to it there is a variable which should contain the value what ever selected in Entity in Page. Now my issue is what settings i should give while making Entity as local variable in BR in EAS.
What should i give in
Type: It has Members, Dimension, String, Integer, Real etc (I tried with Members & Dimension)
Usage Type : It has Saved Selection, Use by Value, Run-time prompt (I tried with all options)
Please give your inputs.
Thanks

Yes i am on version 11.1.2
I am running BR on save of data in Dataform.
I have not set any limits
I have declared a global variable with Member as type and Usage Type as "RTP"
Please let me know if i need to give any default value. My understanding should be no.<
Let me know if i need to click on check box saying "Do not save value entered during validation and launch as default value". My understanding - doesnot matter.<After that the global variable is appearing in BR
Let me know if i need to select checkbox for Hide. My Understanding should be yes<Thanks

Inserting in the UUT_Results table a value that we read from our data base from a local variable

We would like to include in the UUT_Results table a value that we read from our data base into a local variable during the execution of our sequence file. We found that by modifying the configure database options we were able to add a column for this variable, but the local variable was not available to be placed into an expression for that column from the local variables or parameters. Is it possible to do this, and if so, how? Station Globals were available to be included in the expression, however the sequence file may be executed on more than one system which makes the Global unavailable on systems other than the one where the sequence file originated.

You can use the TestStand API to programmatically create global variables at runtime, thus ensuring their existence. For example, you could call Engine.Globals.SetValString("GlobalStringVariableName", PropOption_InsertIfMissing, "variable value")
Of course, if you need to test multiple UUTs in parallel, a single global is not sufficient. In that case you might consider adding the field you need to the UUT datatype in the process model. You could then access the field in your sequence via RunState.Root.Locals.UUT.YourNewField = YourNewValue.
If you also want your sequence to run without using a process model, you must check for the existence of the UUT before accessing it. You could use the expression function: PropertyExists("RunState.Root.Locals.UUT.YourNewFi
eld")

Can I use RunState.ProcessModelClient.Data.Seq[0].Locals.MyVariable in batch model to set variables in a client sequence?

I have used RunState.ProcessModelClient.Data.Seq["MainSequence"].Locals.MyVariable to set Locals variables in client sequences running in a sequential model but it doesn't seem to work with the batch model. Using a breakpoint I found that RunState.ProcessModelClient.Data.Seq is a container listing all of the sequences in the client sequence with the Main Sequence being index 0.
I tried using RunState.ProcessModelClient.Data.Seq[0].Locals.MyVariable to set the variables in my client sequence and it worked fine.
My question: is this method reliable or am I doing something that might cause problems down the road?
Thanks,
Steve

A few issues from doing it this way-
1- You should try an make your process model such that it can handle almost any sequence file thrown at it. MainSequence may not a) contain that local variable and b) be at array element 0. Unless you do some serious error checking then using your process model elsewhere will be a problem because it will throw errors.
2- It all depends on where you are changing that variable at. Remember with the Batch Model that you are dealing with multiple copies of your client sequence. If you change a Locals in the process model which threads does it actually affect?
3- Locals is to be used for the sole purpose of the sequence it resides in. Just like in text based programming where you have a function call:
MyFunc(int x, int y){
string foo;
foo is local and really shouldn't be changed outside of MyFunc. x and y are parameters and should be how data is shared among different sequences in TestStand. There are variables for bigger scopes as well (i.e. StationGlobals, FileGlobals).
I recommend using either a StationGlobal or passing the data as a parameter.
In the end you are the engineer and the flexibility that TestStand gives you makes it so you can do things however you want.
My 2 Cents,
jigg
CTA, CLA
teststandhelp.com
~Will work for kudos and/or BBQ~

Custom class loader and local class accessing local variable

I have written my own class loader to solve a specific problem. It
seemed to work very well, but then I started noticing strange errors in
the log output. Here is an example. Some of the names are in Norwegian,
but they are not important to this discussion. JavaNotis.Oppstart is the
name of my class loader class.
java.lang.ClassFormatError: JavaNotis/SendMeldingDialog$1 (Illegal
variable name " val$indeks")
at java.lang.ClassLoader.defineClass0(Native Method)
at java.lang.ClassLoader.defineClass(ClassLoader.java:502)
at java.lang.ClassLoader.defineClass(ClassLoader.java:431)
at JavaNotis.Oppstart.findClass(Oppstart.java:193)
at java.lang.ClassLoader.loadClass(ClassLoader.java:299)
at java.lang.ClassLoader.loadClass(ClassLoader.java:255)
at java.lang.ClassLoader.loadClassInternal(ClassLoader.java:315)
at JavaNotis.SendMeldingDialog.init(SendMeldingDialog.java:78)
at JavaNotis.SendMeldingDialog.<init>(SendMeldingDialog.java:54)
at JavaNotis.Notistavle.sendMelding(Notistavle.java:542)
at JavaNotis.Notistavle.access$900(Notistavle.java:59)
at JavaNotis.Notistavle$27.actionPerformed(Notistavle.java:427)
JavaNotis/SendMeldingDialog$1 is a local class in the method
JavaNotis.SendMeldingDialog.init, and it's accessing a final local
variable named indeks. The compiler automatically turns this into a
variable in the inner class called val$indeks. But look at the error
message, there is an extra space in front of the variable name.
This error doesn't occur when I don't use my custom class loader and
instead load the classes through the default class loader in the JVM.
Here is my class loading code. Is there something wrong with it?
Again some Norwegian words, but it should still be understandable I hope.
     protected Class findClass(String name) throws ClassNotFoundException
         byte[] b = loadClassData(name);
         return defineClass(name, b, 0, b.length);
     private byte[] loadClassData(String name) throws ClassNotFoundException
         ByteArrayOutputStream ut = null;
         InputStream inn = null;
         try
             JarEntry klasse = arkiv.getJarEntry(name.replace('.', '/')
+ ".class");
             if (klasse == null)
                throw new ClassNotFoundException("Finner ikke klassen "
+ NOTISKLASSE);
             inn = arkiv.getInputStream(klasse);
             ut = new ByteArrayOutputStream(inn.available());
             byte[] kode = new byte[4096];
             int antall = inn.read(kode);
             while (antall > 0)
                 ut.write(kode, 0, antall);
                 antall = inn.read(kode);
             return ut.toByteArray();
         catch (IOException ioe)
             throw new RuntimeException(ioe.getMessage());
         finally
             try
                if (inn != null)
                   inn.close();
                if (ut != null)
                   ut.close();
             catch (IOException ioe)
     }I hope somebody can help. :-)
Regards,
Knut St�re

I'm not quite sure how Java handles local classes defined within a method, but from this example it seems as if the local class isn't loaded until it is actually needed, that is when the method is called, which seems like a good thing to me.
The parent class is already loaded as you can see. It is the loading of the inner class that fails.
But maybe there is something I've forgotten in my loading code? I know in the "early days" you had to do a lot more to load a class, but I think all that is taken care of by the superclass of my classloader now. All I have to do is provide the raw data of the class. Isn't it so?

How to use local variable in member formula?

Hi Expert,
We would like to use a {local variable} in a member formula, so that it can pick up the prompt from webform upon save. Would like to know if it is possible to do? If so, what is the correct syntax?
We attempted to use brace {scenario}, blacket [scenario], blacket with quote ["scenario"]. None of these work.
We are using 11.1.2.1 version, EPMA, Hyperion Planning.
Thanks in advance for your help!

One other pointer. Where did you define your variable? Hopefully you went into Calc Manager and selected <Tools>, <Variables>. From there, navigate to your application or database. Right-click on your application or database and select "New". In the bottom panel, you'll need to select your Variable Type. Make sure you select "Member", and not "Members". You cannot pass "Members" (plural) to a web input form. Also, if you want to define a variable for a custom dimension, you'll need to do it at the database level. Standard dimension member variables can be defined at the app level. (Not sure why this is the case, especially when my custom dimension exists in all plan types . . . . )
Anyway, hope this helps,
- Jake

How can I put the contents of an array into a set of write local variable clusters?

I have a set of controls that a user fills in. The controls are clustered by rows, then combined into a 1D array, then written to a file. When I read the file, I want to feed the array back into the clusters with write local variables. Is there a simple way to handle the numerous local variable clusters so I can stuff the array back into them?
Is there a better way to handle the whole issue of storing and recovering the information?
The rows are steps in a procedure, and each row (cluster) contains numerical as well as string controls.

There are many ways of doing this
You may convert all your data to strings and save as an ascii file (you have to pay attention to the file format, so you can read it properly.
Or you can use the configuration vi's, it may work in your case.
or you may use an invoke node to get all controls of the vi's, save it, then read it, without worrying about any format. I have attached a vi that uses this method.
Attachments:
Temp.vi ‏52 KB

Final attributes and local variables - performance ??

Hi all,
I and a colleague have done some performance testing regarding the use of final attributes and final local variables, e.g.
with final:
public class MyClass {
private final int i;
public final void myMethod() {
    final int j = 5;
    // do something with i and j
}vs. non-final:
public class MyClass {
private int i;
public final void myMethod() {
    int j = 5;
    // do something with i and j
}I couldn't find any speed differences in a small test program, but my colleague did so in his application. Who is right ??
Still, I will have do some formal testing next week and I will post the results.
I'd prefer the version using final anyway because I find it better readable, but the issue I am having is whether I'll spend 2-3 days going through the program making everything final or not.

I made some tests with final arguments to a method: I could not find any difference between final and non-final arguments. code is posted below
import java.io.*;
import java.net.*;
import java.rmi.*;
import java.util.*;
import javax.ejb.*;
import javax.naming.*;
import javax.rmi.PortableRemoteObject;
import junit.framework.*;
import junit.extensions.*;
public final class FinalVariablesTest extends TestCase /* from junit */ {
     * Constructors
    public FinalVariablesTest(String name) {
        super(name);
     * helper methods/classes
    protected void setUp() throws Exception {
        super.setUp();
    protected void tearDown() throws Exception {
        super.tearDown();
     * Test Suite
    public static void main(String[] args) {
     junit.textui.TestRunner.run(suite());
    public static Test suite() {
        return new TestSuite(FinalVariablesTest.class);
     * Test Cases
    /** tests the effect of passing an (final or non-final) int parameter
     to a method which uses the variable in a for loop.
     <p>
     Linux System:
     cat /proc/cpuinfo
     processor       : 0
     vendor_id       : GenuineIntel
     cpu family      : 6
     model           : 8
     model name      : Pentium III (Coppermine)
     stepping        : 1
     cpu MHz         : 501.146
     cache size      : 256 KB
     fdiv_bug        : no
     hlt_bug         : no
     sep_bug         : no
     f00f_bug        : no
     coma_bug        : no
     fpu             : yes
     fpu_exception   : yes
     cpuid level     : 2
     wp              : yes
     flags           : fpu vme de pse tsc msr pae mce cx8 sep mtrr pge mca cmov pat pse36 mmx fxsr xmm
     bogomips        : 999.42
     </pre>
     <p>
     Results:
     <pre>
     java version "1.4.0"
     Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.0-b92)
     Java HotSpot(TM) Client VM (build 1.4.0-b92, mixed mode)
     final     non-final
     498     500
     491     494
     491     493
     491     494
     534     494
     492     494
     491     494
     492     493
     491     494
     495     494
     4966     4944 (Totals)
     </pre>
    public final void testIntParametersToForLoop() {
     final int RUNS = 10;
     final int INNER = 1000000;
     final int OUTER = 10;
     System.out.println("-----------------------");
     System.out.println("testIntParametersToForLoop");
     for(int i=0; i<RUNS; i++) {
         outerFinalIntParametersToForLoop(INNER, OUTER);
         outerNonFinalIntParametersToForLoop(INNER, OUTER);
    private final void outerFinalIntParametersToForLoop(final int INNER,
                                   final int OUTER) {
     // with final var in for loop
     long start0 = System.currentTimeMillis();
     for(int i=0; i<OUTER; i++) {
         innerFinalIntParametersToForLoop(INNER * i);
     long end0 = System.currentTimeMillis();
     System.out.println("      final:       " +
                  ( end0 - start0 ) + " milliseconds");
    private final void outerNonFinalIntParametersToForLoop(final int INNER,
                                      final int OUTER) {
     // with non-final var in for loop
     long start1 = System.currentTimeMillis();
     for(int i=0; i<OUTER; i++) {
         innerNonFinalIntParametersToForLoop(INNER * i);
     long end1 = System.currentTimeMillis();
     System.out.println(" non final: " +
                  ( end1 - start1 ) + "       milliseconds");
    private final void innerFinalIntParametersToForLoop(final int INNER) {
     for(int i=0; i<INNER; i++) {
         int testVar = i * INNER;
    private final void innerNonFinalIntParametersToForLoop(int loops) {
     for(int i=0; i<loops; i++) {
         int testVar = i * loops;

Does a large amount of "local variable" within a vi cause problems?

Hello -- I am using Labview 7.0 and have a vi where I am going to end up using about 100 "local variables" for indicators that are displaying data. Since I have a lot of true / false case structures I decided to use these local variables instead of having a mess of wiring all over the vi block diagram. It is intimidating though and I was wondering if it causes slowness in vi's?

I actually only have about 8 indicators -- but I am in the process of creating control algorithms so I have to use many case structures if I want to keep the code all within Labview (simplicity). I am simply referencing the 8 or so indicator values many many times (100 X) I am attaching a copy of the vi like you asked but just take a look at the right hand side of the while loop on the right -- it shows an example of an algorithm I created today that has to reference the data from sensors. I would love to know what a better way to do this. Thanks!
Attachments:
ucb drill control daq 1_7.vi ‏666 KB

Private data members vs. local variables

Similar Messages

Maybe you are looking for