DIAdem Idea Exchange

Community Browser
cancel
Showing results for 
Search instead for 
Did you mean: 
Post an idea

It would be nice to have the Report panel included stacked bar charts.  When looking at a number of inputs to a process while changing those inputs, it's helpful to easily see the relative contribution of each input to the total output. 

 

http://forums.ni.com/t5/DIAdem/Bar-Chart-with-moret-hen-1-Channel/m-p/3152564/highlight/true#M20550

The subject already says everything. I assume it takes just a little effort to implement, but will make scripting more ergonomic.

I created a service request asking for a way to do this after reading the documentation and the person that answered the request confirmed that there's currently no way to do this. He also recomended that I posted the information here, in order to include this feature in future versions of DIAdem / DataPlugins.

 

The file path I want my dataplugins to obtain is the path of the dataplugin’s own script file, including the filename. For example: if my dataplugin is named “ABCD.vbs”, then I would like that dataplugin to find out its own absolute path, which may be something like “C:\Dataplugins\ABCD.vbs”. Having a way to do that would save us the trouble of hardcoding it and keeping track of the path if it changes for every single dataplugin.

For scripts that run on the Windows Script Host we can use Wscript.ScriptFullName to obtain it and for scripts that are run in DIAdem’s scripting context we can concatenate the DIAdem variables CurrentScriptPath and  CurrentScriptName. What I want is to have that same functionality but for dataplugins.

Could there be a way to highlight or favorite lines of code that are interesting or problematic? I often find myself searching for the same couple of lines for use, reference, or problem solving, and they would be much easier to find if there were a way to highlight them in some way.

Hey everyone,

There should be an integrated feature in DIAdem to create Boxplots. You can already find the min Q1 median Q3 and max in the analysis portion. I have not been able to find a good way to make a boxplot in DIAdem, and it's a very clunky process in excell. I've had to use DIAdem to process data then export the interested channels to excel to make acutal boxplots, which still requires a lot of manipulation.

BoxPlot Data.PNG

 

To do this in excel you create a modified boxplot based on the difference in the five number summary. These numbers are put into a stacked boxplot, then the settings are played with to give the finalized look of a box plot.

Differences.PNG 

The Differences plotted as a stacked bar plot looks like the following with borders on and the bottom and top sections made invisible with error bars to mark the min and max values

boxplot.PNG

There should be a way to do a similar process, or even an automatic box plot creator in DIAdem.

 

Let me know what you think,

Sadie

The new Event Search in 2015 is nice. But it does miss one feature. The need to allow output of the original value if TRUE, or better yet, a function calculation.  Right now, the Event Search only allows a static numerical value to be entered when the condition is TRUE, then another function call like Multiple from the original channel and the Event Search channel has to be performed to create a "filtered" channel of the original values.  Having functions for both the TRUE and FALSE conditions could hep with filtering and charting. a 0 result for false isn't always the best option.

I have been doing some data analysis in DIAdem; I have been taking some statistics on some data - working out variance and standard deviation.  I asked NI Support team whether 'Population' or 'Sample' weighting is used, as it was not mentioned in the help file. I understand that by default Variance and Standard Deviation are calculated using the 'Sample' rather than 'Population' formula. My company uses Population to work out some of the key performance data of the products we make.  At the moment I don't see a way of selecting which formula to Descriptive statistics.PNGuse so a little drop down next to the checkboxes so I can pick the type would be great.  For now I'm going to have to develop some VBScript or start using excel to do my data analysis.  As a footnote to this idea - may I also illustrate that calculating standard deviation directly feeds into process capability (Cp and Cpk).  Sample weighting is a little less friendly than Population weighting, so unwitting process control engineers may be finding their processes a little more (I think!) capable than what they truly are.

 

Regards,

 

Sam

 

Hi,

 

It'd be nice to be able to clear the data portal without having to go to the NAVIGATION area each time. Adding the ability to clear the data Portal from any screen, especially when scripting, would be much more convenient. 

 

Regards,

Sadie

There is obviously a problem converting time Strings like 1.5.2015 correctly to time data and back. Everything works well as long as Day and Month have really both two digits. As long as this is not fullfilled, the conversion is wrong when the TTR is then used like this:

 

D2AXISXBEGIN=TTR(ZeitMin_,"#DD.MM.YYYY")

 

Then 2000 years missing and the date arriving the report is 1.5.0015 what is in fact slightly different. I off course programmed a workaround filling up the "0" in front of a one digit day and month, but that shout be done by the TTR correctly.

It would be very useful to be able to copy data directly from a View 2D axis chart/graph to the clip board.

This action is similar to using the channel table to copy a selection of data and being able to past that data to the next empty column of the table or pasting it directly into the portal to create a new channel.

 

Currently I can zoom into a small region of the data in the View graph.  Based on the  position of X-cursor1 and X-cursor2 I can find the data in the table and copy it to the clip board.  Then create new channels by pasting into empty columns or pasting into the portal.

But this is very time consuming and cumbersome.  

 

Some time ago I created a script that would copy the data to new channels automatically, but if I can copy the data to the clip board I can paste it into other programs like, please forgive me, Excel for customers who don't have Diadem.

 

Excel does have the TDM importer add-on, but then it is really hard to search for data.

 

 

 

The idea is to provide a "Channel concatenation" button in the default Analysis panel > Channel Functions.

After clicking this button, a dialog box should open, where the various Channels of different Channel Groups can be selected for the concatenation process. In the same dialog box, some options for the channel concatenation could be selected as well.

In this moment, some example scripts can be found in the discussion board for the concatenation of N Channels in M Groups. However, these scripts are quite cumbersome, not easy to customize, and providing too many optional features that are aiming to address shortcomings in the time or data channels to be concatenated.
It is a good practice, however, to correct first the issues in your Time and Data channels that prevent them of being concatenated properly.

The channel concatenation function would be a very useful feature, especially for engineers working with big datasets, that are increasingly recorded in contemporary data acquisition.
Typical examples of big datasets are vehicle CAN network data, recorded over an extended period of time (weeks, months).
Usually, these data are divided into multiple data files, which then need to be concatenated afterwards in order to cover a selected recording period, prior to further analysis of the data.

 

Currently the data type of group properties is automatically set, when two colums are used, one for "Name" and one for "Value". (see picture, right side)

I recommend to enable the possibility to manually set the data type for every individual entry, regardless which column is chosen for Name and/or Value.Bild1.png

With large datafiles and hence,many points in a x-y-graph it takes very long for diadem to update all curves in a report. It is really annoying if you need to change the layout on many report pages since the layout is updated every time you are changing it a bit...

 

Is it possible to suppress the automatic update of the whole page? If necessary, the user can still update the report page manually by using the update-button.

Add the ability to decode relative paths for video files within View and TDV files AND make it the default setting.  Adding an option box to the "Select Video File" dialog to allow absolute or relative path decode.  Also have the system recognize a relative path (i.e. ".\video.avi").  This may require a new TDV-Path variable.  This allows moving large video/data views to networked based folders allowing other engineers/customers to review/manipulate the data.

 

 

Hello,

 

Because it is only possible to play a driven route including graphs in the DIAdem VIEW and the cursor is hardly visible in this view it would be a good feature if it is possible to change the cursor style/cursor line thickness. See the attached video for an example of the hardly visible cursor.

 

With kind regards,

Stijn

When I export a plot in a DIAdem report I lose the transparency that I have associated with any coloured background segments in my plots.  It would be helpful if the transparency was maintained as it helps to show overlapping areas.

 

As a side note, I find that when I zoom out on the axis of a plot the background segments disappear when they get narrow, but still wide enough to show on screen.  This can be confusing as there is no indication it is happening.  Perhaps background segments could always be drawn at least 1 pixel wide.

 

Thank you,

 

Martin

Hello,

 

I checked the DIAdem hepl and saw that the time value converted in text only allows 4 digits for the fraction of a second.

But how can you display datetime in text for more precise values.

 

For example, if i acquire on a real time target, analog inputs data at 1MHz, all samples are more precisely time stamped than the best text value that you can display.

 

It would be very convenient to display time stamp like that : 06/19/2013 15:37:50.12345678

instead of that : 06/19/2013 15:37:50.0005

 

Thanks  !

Best Regards

CFOE

Not having the DataFinder index anything except NI's folders by default on install would be extremely convenient.  Optionally, this could also be a setting chosen on install.

 

Each of our users who installs the application immediately has every MS Excel file from their 'My Documents' folder indexed, which wouldn't really hurt anything except that most of these MS Excel files aren't true 'data' files.  In my experience, most Excel files rarely store anything as simple as what could be stored in text-tab delimited format, so aren't easily or usefully indexed, so seems to confuse the DataFinder as it is unable to close them automatically once opened.

 

The results is dozens to hundreds of Excel.exe processes running in the background, until the machine runs out of memory or files, so we have had to make a point to disable all datafinder indexing on each and every install.  We've scripted this process into our standard configuration, but if the user doesn't contact us right away, they suffer greatly from performance issues and blame DIAdem (rightly so).

 

I suspect that anyone who wants to use the DataFinder would want to do some manual configuration to specific folders, dataplugins, times of day to index, etc. which would make requiring setup perfectly acceptable to users who want to use it (and not waste resources for those that don't).

 

-Josh

global Variables do not have the "Auto Complete" Funktion when you write them in script.

When editing a Script. The path for the saving files  defaults to the last opened file, instead of remembering the path for each file being editted.  This causes files to be saved in the directory of the last file opened, which many times results in the wrong saving location. This is especially an issue when have large number of script files in different directories.

 

I would prefer that each script file that was being edited would keep it path separate and stored so that it always saves the files to the location that it was opened from.