Quantcast
Channel: OSIsoft Users Community
Viewing all 1120 articles
Browse latest View live
↧

Time Syntax

July 19, 2013, 1:04 am
$
0
0

Hi, it's me again.

 

This time I've a question to the PI Time Syntax.

 

In my case I like to compare a current tag value with the value of now before 30s.

 

So the PI Calculation is:

 

'Tag_Name' - PrevVal ('Tag_Name','*-30s')=0.

 

Is it true (1) -> Alarm

or False (0) -> Ok.

 

But my problem is... In PI ProcessBook it is possible to select a time range in past. Which code is used to use this bygone time as now?

 

So for example. Now is 9:55 am. I go back to the time range 6:55 am to 7:55 am. The current time in the PI Calculation should be now 7:55 am.

 

I hope you guys get my problem and you have a good tip for me.

 

Thanks a lot.

 

 

↧
Search

Datasets slowing down displays

July 19, 2013, 4:02 am
$
0
0

I use a lot of datasets.  More than a few on a graphic makes the graphic so slow it is unusable.  I slowed the calculation intervals of 30 minutes but that didn't help.  Any suggestions?  

↧
↧

Here's why there's Twitter stuff on the community front page.

July 19, 2013, 11:54 am
$
0
0

Maybe you have noticed the twitter widget on the lower right of the front page of the community. This isn't a feed that comes from OSIsoft, although we do have some twitter usernames that send tweets regularly, like @OSIsoft_News (marketing), @OSIsoftPSS (Partner solutions), @OSIsoftUC (users conference), and @OSIsoftSupport (technical support). 

 

Instead, it is a simple search result of all tweets that contain the word "OSIsoft." And that is why you will see things there that may be off topic. This is admittedly a scattershot approach to getting value from Twitter, but it works so often that there are more reasons to do it than not to do it. If you go through the list you will find many things you are not interested in, such as recruiters looking for PI System Managers, or people who happen to be visiting us today (like the job candidate who tweeted "Interviewing at OSIsoft today!" not knowing that it would appear on our front page).

 

But occasionally you will find things that you otherwise might miss. For example last week you would have found prolific blogger Oliver Yu's poston how your PI System can be used for more than just compliance. And in a tweet from @Agrion I learned about this streaming panel discussion on Harnessing Data for Sustainability Management and Driving Successful Programs that features OSIsoft's Kristen Kelly as one of the panelists.

 

So here are some typical tweets that appear there today:

 

twitter_01.png

 

The first one is from @cpapenfuss, who is OSIsoft employee Christoph Papenfuss. The #FF stands for "Follow Friday" and is a tradition on twitter to send a tweet with the handles of people you suggest others "follow" (i.e. subscribe). So if you like the work Christoph has done in his Performance Ideas blog (which you should check out if you are interested in data visualization writings of Stephen Few or Edward Tufte), then you may want to follow on Twitter the folks he recommends in his #FF tweet.

 

The second one is a job posting. We don't endorse them of course, and we don't ban them.

 

The third is a tweet from Lonnie Bowling ("Lonnie" in this user community), who is sending an update about a series of screencasts he is considering doing about PI System programming. If you click on @LonnieBowling and go to his full profile--no Twitter account required to do that--then you'll see earlier tweets that describe this series and has a link to an OSIsoft vCampus blog post that describes the series. 

 

So please consider looking through the tweets daily. If a lot of your job involves working with the PI System I think on a regular basis you will be able to find gems of useful information there.

↧

PI Datalink - data filtering

July 22, 2013, 7:56 am
$
0
0

Hi,

 

I want to retrieve a list of PI data points without the same data value for example:

 

Original list:

Timestamp                    Value

-------------------------     -----------------------

01-JUL-13 00:00:01     1

01-JUL-13 00:00:03     1

01-JUL-13 00:00:04     1

01-JUL-13 00:00:05     2

01-JUL-13 00:00:06     1

01-JUL-13 00:00:07     3

01-JUL-13 00:00:08     1

01-JUL-13 00:00:09     1

 

Expected result: (get the changed data points only)

Timestamp                    Value

-------------------------     -----------------------

01-JUL-13 00:00:01     1

01-JUL-13 00:00:05     2

01-JUL-13 00:00:06     1

01-JUL-13 00:00:07     3

01-JUL-13 00:00:08     1

 

Can I achieve the above result through PI datalink? or VBA?

 

Thanks in advance for any help.

 

 

 

↧

An approach to querying the pi system with Business Objects

July 24, 2013, 4:06 pm
↧
↧

ProcessBook offline mode

July 25, 2013, 8:53 am
$
0
0

Hi guys,

 

Is there a way to smoothly work on PB display in the offline mode? I know I can disconnect server connection, but the display hangs up if I do that as it keeps attempting to connect.

 

Akshay

↧

store "spectrum" of data in one PI point

July 26, 2013, 7:58 am
$
0
0

I'm sorry to create a second topic regarding this; the previous topic is marked as 'answered', but maybe I did not gave enough information because the answers do not solve my specific problem.

 

What I want to do is store in one PI point an array of values of a spectrum and preferably display this spectrum using an OSISOFT tool: so an array of values each timestamp instead of only one value each timestamp in one PI point. This is because I receive a spectrum (array) from my PAT analyser instead of one value each timestamp. 

 

F.e. instead at 7/26/2013 16:30:00 -> value: 14.3, I get at: 7/26/2013 16:30:00 -> 10 values:14.3, 20.8, 31.7, 12.4, 19.2, 20.3, 5.0, 4.0, 3.2, 0.1.

 

One possibility is to use several PI points (0%, 10%, 20%, ...100%) for a rough spectrum each timestamp, but this gives me not enough information or else I have to create too much PI points: the problem is that I have a spectrum that has multiple peaks. For sufficient distinction I need about 100 values in the array or about 100 PI points.

 

Another possibility is using one string point, and put all the values of the array in this string in CSV format, but then I do not have an easy way to display the spectrum. Maybe in XL and some VBA for parsing the string or with a PB add-in?.

 

Any other possibilities?

 

Thanks, Johan.

↧

Cursors on all trends could be a wonderful thing

July 27, 2013, 1:15 pm
$
0
0

I want some of my trend displays to open with a cursor on each trend.  Of course I plagiarized some nifty code and adapted it, and it works great when I run it.  But when executed via the displays open event all the cursors show "no data" I think because it places the cursors before the trends have data.  What should I do?  !!  

 

Here is my code:

 

 

Private Sub Display_Open()
 
    Call SetTrendCursors
 
End Sub
 
    Function SetTrendCursors()
 
    Dim TrendCnt As Integer
    Dim Ndx As Integer
    Dim SymType As Long
    Dim NumCursors As Integer
    Dim ThisCursor As Integer
       TrendCnt = ThisDisplay.Symbols.Count        ' Find out how many symbols are on the display page
    ' Walk down the list of Symbols and if the Symbol is a Trend, Set a Cursor
    For Ndx = 1 To TrendCnt
        SymType = ThisDisplay.Symbols.Item(Ndx).Type
        If SymType = pbSYMBOLTYPE.pbSymbolTrend Then
            
            'Remove All Cursors
            While ThisDisplay.Symbols.Item(Ndx).CursorCount > 0
                ThisDisplay.Symbols.Item(Ndx).RemoveCursor (1) ' ThisDisplay.Symbols.Item(Ndx).CurrentCursor
            Wend
            Call ThisDisplay.Symbols.Item(Ndx).AddCursor
        End If
    Next
End Function
 
↧
Search

New videos on how to import PI System data into PowerPivot, the Excel Add-in BI tool

July 28, 2013, 1:20 am
$
0
0

The Learning team has published a new playlist on the OSIsoftLearning YouTube channel on PowerPivot and PI System Data.  Users are integrating their data with some of the Microsoft Business Intellegence tools and these new videos teach how to use PowerPivot with the PI System. 

 

Here is the first in the series. 

 

 
The video series is presented in steps that are broken into the numbered sections below. You can jump directly to each numbered topic from links on the descriptions in YouTube. 
 
Step 1 Prepare the PI AF Structure
  1. Choosing an AF Server
  2. Organizing a PI AF Server
  3. Importance of Using Templates
 
Step 2 Create Transpose Functions with SQL Commander
  1. Load the PI AF Structure with PI SQL Commander
  2. Resources for building detailed queries
  3. SELECT queries for PowerPivot can be general
 
Step 3 Create Views with SQL Commander
  1. Start from default View of a Transpose Function
  2. Remove unneeded columns from default View
  3. Q: What are Views? A: Named queries
 
Step 4 Import Views with PI OLEDB to Excel
  1. Starting with a View from SQL Commander, Open PowerPivot
  2. Import to PowerPivot: Other Data Sources: OLEDB/ODBC
  3. Build a PI OLEDB connection string
  4. Enter 'SSPI' for Integrated Security 
  5. Specify PI System & Test Connection
  6. Select PI AF Database
  7. Review Connection String
  8. Select Tables and Views to Import from PI to PowerPivot
  9. PI Data is now in PowerPivot! It can be refreshed anytime

The playlist can be found here.

↧
↧

XYPlot, XYTags and XYTag

July 29, 2013, 1:47 am
$
0
0

Goodmorning everyone,

 

After hours of tests, I come to you with the hope that you can help me to find a solution to my small problem... :)

I have created a XYPlot on my ProcessBook.

Then, I would like to access to the tags of the XYPlot using VBA language.

 

By looking at the help I understand that I should use a XYTags collection and the function XYTags.add, but it does not work, and I also don't know how to link the XYTags collection to the XYPlot. So I'm lost....

 

I do the same "exercise" very often on standard trends, adding a tag to the trend using the function:

 

ThisDisplay.mytrend.addtrace tag

 

And it works very well.

But for XYPlot trends, I cannot find the way to proceed...

 

Any idea ?

 

Thank you in advance !

 

Vincent

↧

Importing Contacts into PI AF Notification System

July 29, 2013, 6:42 am
$
0
0

I am fairly new to the PI AF system and I am currently attempting to set up an email notification system through the software yet I have run into an issue.  The PI AF system does not contain any of my companies emails loaded into the system so that I can send out these email notifications.  Is there a possible way that I could import these emails/ company email directory into the PI AF system?

 

Thank you,

 

Cody Balzen

↧

Steam tables in PI

July 30, 2013, 1:15 am
$
0
0

Hi

I know that PI Performace calculations includes the ability to get Steam table data. Ethalpy, entropy, Saturated steam etc etc.

could one of the more knowledgable person assist me with getting the correct formats

I have managed by Trial and error to get the following working

stmSi_PsatT this seems to give steam saturation pressure with temerature as is basis. .

for example

StmSI_PsatT(100)

will return a value of 101.325275 the result must be in Kn/M3 as for absloute pressure in bar it should be 1.013 for mbar 1013 etc.

The value in brackets can be any suitable Tag from the server

 

does anyone know what the units are for the stmSI expressions are, and howmany variant of stmSI/stmENG are there?

google is not very helpfull today

any help would be very welcome.

Ian Fuller

 

Data is a beautiful thing, the more varied the better.

↧

SVG file embedding jpg images problem in deployment on sharepoint

July 30, 2013, 10:26 pm
$
0
0

Hello everybody,

 

We have a processbook display embedded with JPEG images, we are able to retain the image when saved as .svg. When I deploy .svg file on sharepoint it is not visible. Is there something missing here?
 

Regards,

ilyas

↧
↧

Create a constant value tag

July 31, 2013, 12:55 pm
$
0
0

What is the easiest way to create a constant value tag?

↧

PI Event Frames + PI DataLink = Events in Excel (a.k.a. Awesome)!

August 1, 2013, 7:27 am
$
0
0

PI Event Frames + PI DataLink CTP is Almost Here!!!

 

Hey everyone, if you're interested in trying out the upcoming PI Event Frames functionality in PI DataLink (and giving us your feedback before release), please read the following blog from my fellow Product Manager Matt Ziegler and sign up for the community technology preview (CTP) [beta].  We should have the CTP available sometime in September!

 

http://community.osisoft.com/index.php?/page/categories.html/_/matt-ziegler-s-blog/pi-betalink-2013-r2-register-now-to-be-the-first-to-get-the-new-bits-r38

 

Also, make sure you poke fun at Matt for naming his post PI Betalink instead of PI DataLink!  or knowing Matt, maybe that was on purpose ... hmmmm   :)

 

Here's just one of the 'event-enabled' PI DataLink reports we showed at UC2013!

PIEF_DL1.jpg

 

And here's a recent screen shot of the PI DataLink team's great work!

PIEF_DL2.jpg

 

Thanks,

 

Todd Brown

PI Event Frames Product Manager

↧
Search

Time Filtering Compressed Data

August 1, 2013, 7:56 am
$
0
0

The back story: We have an analyzer which gives a new output every 20 minutes and holds that output for 20 minutes until it processes the next sample and then changes the output to the new value. So if we look at the compressed data, what we should see is Date/Time value, Date/TIme + 20 min, value, etc.


Now what is happening is that somewhere in the system and I assume that it is in the PI data sampling section, PI is adding a pseudo point one minute before the actual point. As a user, I have been unable to locate anyone at our work who can explain why the extra data point is being added and sofar no one has the time to help me resolve this issue, so I am forced to manipulate the data and delete the extra point myself.


My theory is that with a stair stepped input, PI is adding a data point so that it is able to trend the last time period (20 min) worth of data. I'm guessing that there is a switch setting somewhere to tell PI not to do this but unfortunately have no control over the creation of the data. This bogus point is generally close to the actual value but sometimes is way off. And I know that it is somewhere in our system from the RTU to being stored in PI as I have checked the actual outputs from the analyzer to be sure it is not putting out two values.


My idea is this, using Compressed Data to retrieve the TagName's values, I want to filter the data selected to ignore a reading where the NextEvent is less than 2 minutes from the current event being processed. But doing this in PE, I need to know what the current time of the data is so I can check its time against the time of the next event.


If you are retrieving data over say a month and grabbing every reading, I need to be able to deterime the time of the current reading being processed?


Anyone got an idea on how to do this? It's no problem doing it programatically but I would like to be able to display the actual data in ProcessBook.

 

Below is a sample of the data and I have marked the extra readings.

 

01-Jul-13 00:02:35 55.6636467 FALSE
01-Jul-13 00:03:35 52.35025024 
01-Jul-13 00:23:35 52.35025024 FALSE
01-Jul-13 00:24:35 52.92341995 
01-Jul-13 00:42:35 52.92341995 FALSE
01-Jul-13 00:43:35 53.27281952 
01-Jul-13 01:02:36 53.26889038 FALSE
01-Jul-13 01:03:35 51.5258255 
01-Jul-13 01:22:35 51.5258255 FALSE
01-Jul-13 01:23:35 50.09289551 
01-Jul-13 01:42:35 50.09682465 FALSE
01-Jul-13 01:43:35 49.99868011 
01-Jul-13 02:02:35 49.99868011 FALSE


 

TFG

↧

FLASH ANIMATED BUTTONS

August 2, 2013, 2:44 am
$
0
0

I'm trying to enrich the users browsing experience with PBook by using flash animated buttons, these days you can build your buttons online, generate the swf file and simply add it as a control.

The problems starts when you click on the actual button it opens the pdi file in your web browsers and not in pbook, any idea how this can be solved

The current URL for this test is file:///C:\test.pdi

 

Any ideas how I can open the pdi file in processbook using a flash control embedded in processbook?

↧
↧

How to change multi-state in batch?

August 3, 2013, 8:35 am
$
0
0

Hi! Im new in this site.  I'm a user of processbook and wondering on how i can change multi-state in batch? If i have 100 values with 3 multi-states each and decided to have 4 multi-states how can i do it.  When I select all the values and click the multi-state button and changed it, all the values now have the same tag.  Thank you in advance.

↧

PI Notifications Event End Time

August 7, 2013, 8:24 am
$
0
0

I have several PI Notifications set up to send me an email in the event process conditions exceed certain values and need to be reported for environmental reasons.  I noticed that you could set up the PI Notifications to send an email at the start of an event and include the end time of an event within that notification email.  Upon receiving one of these emails though I noticed that the start time was correctly stated however the end time was incorrectly stated on the email as,

 

End Time: 1/1/1970 12:00:00 AM Central Standard Time(GMT-06:00:00)

 

My question is, is there a way to fix this to display an events end time, or is it possible to set up a secondary email to be sent once the event is over?

 

Any help on this issue would be greatly appreciated.

 

Thank you,

 

Cody 

↧

PI Expression - convert PI data value

August 7, 2013, 12:38 pm
$
0
0

Hi,

 

I have a PI tag which sometimes has the following compressed data:

on

shutdown

on

shutdown

on

off

shutdown

off

shutdown

 

I need a expression to convert all shutdown values to the previous value and calcuate the running hours for the value "on" or "off".

Is it possible to use PI expression to achieve this?

 

Thanks a lot.

↧
Viewing all 1120 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>