This discussion is dedicated to help learners of the Cognite Data Fusion Fundamentals learning path succeed. If you’re struggling with the exercises in this learning path, try the tips & tricks below or post a comment with the challenge you’re facing. You can also post your own tips and respond to fellow learners’ questions. Cognite Academy’s instructors are also here to help.
Page 1 / 5
@Charlotte Waese good you figured out!
I found the problem. I accidentally deleted the “ : “ when creating the externalID for the time series. Thus the data points could not be matched. When I put the “ : “ back in to the SQL it worked!
Hi all, I am just running through the Working With CDF: Integrate course and am facing a problem with inserting the data points into the time series.
I am following all the steps and at the end I should be able to see data points in the IFSDB time series data set. However, there are no data points in any of the time series.
The run states that 4,296 rows read - 15,428 timeseries read. So there should be data in IFSDB.values (the reference in the SQL query)
I also deleted the transformation once and did it all again, as well as changed the time interval to 5 years in case the data is older. Still I can see no data.
Has someone an idea I can try?
FYI here the SQL query from the training
SELECT
concat('FirstnameBirthyear:',dp.sensor) AS externalId,
cast(time_stamp/1000 as timestamp) AS timestamp,
cast(value AS double)
FROM IFSDB.values AS dp,
--selecting from _cdf.timeseries means we select from the timeseries we ingested to CDF earlier. We do this to make sure all the time series we try to add data points to actually exist
_cdf.timeseries AS ts WHERE CONCAT('FirstnameBirthyear:',dp.sensor) = ts.externalId
I’m having issues with Transformation of the Raw data provided. Running SQL Script:
SELECT concat('Brad1971:',loc) as externalId, IF(parent_loc='' OR parent_loc IS NULL, '', concat('Brad1971:',parent_loc)) AS parentExternalId,
CAST(lastUpdatedTime AS STRING) AS name,
to_metadata(*) AS metadata,
description AS description,
2147974692941503 AS dataSetId
FROM `Brad1971`.assets`
Is successful preview, but when I execute the query with
It gives an error:
I’ve tried it multiple times and deleted table (starting over) and cleared Raw table to start over as well as Target Table. No Luck. In reviewing the Raw Data Table, I don’t see is missing or bad ID. i could trace the parentID’s all the way up the tree. Please advise.
Sorry...I had selected Asset and not Asset Hierarchy….all good.
Hi @Bjarte Håtveit
I experienced the same myself this morning and I will have to talk to the engineering team about it.
Hi!
I am in Fundamental course and I am working on the exercise “Working With CDF: Contextualize”. The challenge I face is in the part “Contextualize P&IDs”.
My dataset with my assets does not show up in the drop-down - what could be the mistake I have made? (I have done my best to follow the training material)
Best regards,
Bjarte
Thank you for sharing your solutions with the community, @Hannah Håland. This is incredibly valuable for other learners of Cognite Academy. Good luck with gaining your CDF Fundamentals certificate!
Hei,
Yet again, I fix my own problem!
I close the browser and re-sign in and my dataset magically appears :)
Hi,
I am going through the module Working With CDF: Contextualize, and I have loaded the P&ID. Now I am trying to link it to my Aveva assets. But I cannot find my dataset in the Data set drop down menu.
I have checked in the Data Catalog and the data set does exist in CDF.
How do I fix this?
I’m glad you found out @Hannah Håland! Let us know if there is anything else we can help you with!
Actually, I found the mistake from a previous post.
The transformation was wrong in the first place - dataSetId was supposed to be the actual number ID that was recorded down earlier :D
Hi,
I am going through the Working With CDF: Integrate module, and have transformed the data and it looks like the 36 assets have been created based on this output:
But I cannot see the assets in the Data Catalogue:
Do you know what’s wrong?
Thanks Sofie! That did the trick
Hi @Salma Ghafoor
It seems to be due to the configuration of the transformation. The Target resource type needs to be “Asset Hierarchy” instead of “Assets”. The Assets target resource type does not recognize the parent external id attribute. You can change it from the transformations UI by clicking on the dropdown next to “Create or Update Assets”
See the pictures below.
I’ve uploaded assets.csv into a dataset and trying to run the SQL script to create the assets but am getting the error Request with id 87b97f14-7c9b-94b9-9e1f-129e2ed18250 to https://api.cognitedata.com/api/v1/projects/cdf-fundamentals/assets failed with status 400: Reference to unknown parent with externalId Salma:23-PT-92535.
The query runs fine in Preview mode but fails to run. Funny thing is it processes a different amount of rows each time it runs.
Here’s the query I’m using:
SELECT concat('Salma:',loc) as externalId, IF(parent_loc='' OR parent_loc IS NULL, '', concat('Salma:',parent_loc)) AS parentExternalId, CAST(lastUpdatedTime AS STRING) AS name, to_metadata(*) AS metadata, description AS description, 6295592107437051 AS dataSetId FROM `Salma`.assets
Any ideas what’s going wrong?
Hi @Sanyogita Wable ,
I've also tested this, and it appears to be functioning correctly.
Could you please retry in an incognito window to ensure that any potential browser cache-related issues are ruled out?
Regards,
Kanchana
Hi @Sanyogita Wable! I just checked and it works for me. Maybe you @Dilini Fernando can take a look as well?
Enter cognite-learn as a domain name. after enter domain name the error -An error occurred when looking up the details for this organization
Hi again @DE KOCK Kyle
Thank for your reply! Good to hear that you finished! I will give the feedback to the teams responsible for the tools you mention!
Thank you!
Hi @Sofie Haug,
I archived all of my data as I was able to complete the course and do the test after I deleted and recreated my timeseries as mentioned at the beginning of my post. Once that was complete I was able to perform the contextualization with Quick Match but as I show in my post there were some problems with the GUI not showing the entities that match the search parameters of the quick match (this was while my data was not archived). Interestingly enough this was only the case with the 12 timeseries and not the 36 assets.
Another thing I should also mention is that, I was unable to find any data in the graphs after I recreated the Timeseries transformation regardless of the time range (I know the course mentions November 2020 - 2021). This was also while my data was not archived.
Hi @DE KOCK Kyle ! Thanks for your feedback! When I try to take a look at the data, I find your data sets as archived. Did you do this after you tried the matching? Some of the search fields filter away archived data sets.
I see that you have the assets, time series and events data. Your time series data is also contextualized to the relating assets. Did you manage with the Match resources? Or is this contextualization from using transformations?
Hello All,
I was going through the training yesterday and got all the way to the end of Working with CDF: Contextualize and the system got stuck trying to create the interactive PID. Fast forward to this morning and I re-did creating the interactive engineering diagram and it worked.
I then moved on to the next course Working with CDF: Consume and while I was trying to find my timeseries I noticed that my Kyle1995-IFSDB data catalog no longer had any time series associated with it (see below). I then tried to do the data transform for the timeseries data and it tells me that it has updated 12 timeseries but I still cannot find those timeseries in my data catalog anywhere(see below).
I should also note that I did do the optional contextualization exercise of linking the timeseries to assets. I then also ran the contextualization using the entity matchers.
What can I do to get my timeseries data back so that I can continue on with the training?
The only was able to get the timeseries to show in my data catalog again was to delete the FirstnameBirthyear-Timeseries Data Transformation and re-do it.
I also found a bug in the Contextualize > Match resources to assets > Quick Match page where it doesn’t show the entities that have been found (see below in red) despite there being results (see below in cyan).
This issue also seems to result in not being able to create additional fields when configuring the model, I guess due to the fact that the system doesn’t seem to think that anything was found in the timeseries step (see below in red).
That being said it still appears to run based on what is actually in the system and not what the GUI is showing.
Hello All,
I was going through the training yesterday and got all the way to the end of Working with CDF: Contextualize and the system got stuck trying to create the interactive PID. Fast forward to this morning and I re-did creating the interactive engineering diagram and it worked.
I then moved on to the next course Working with CDF: Consume and while I was trying to find my timeseries I noticed that my Kyle1995-IFSDB data catalog no longer had any time series associated with it (see below). I then tried to do the data transform for the timeseries data and it tells me that it has updated 12 timeseries but I still cannot find those timeseries in my data catalog anywhere(see below).
I should also note that I did do the optional contextualization exercise of linking the timeseries to assets. I then also ran the contextualization using the entity matchers.
What can I do to get my timeseries data back so that I can continue on with the training?
Hello,
For the Data points step “Transform RAW to Data points” I received the error “Query returned no results. The query is syntactically correct, but returned 0 rows of data.”
It seems the query does not filter for dataSetId and returns no matches. I changed the query to solve the issue on my end, here are the changes:
SELECT concat('FirstnameBirthyear:',dp.sensor) AS externalId, cast(time_stamp/1000 as timestamp) AS timestamp, cast(value AS double) FROM IFSDB.values AS dp JOIN _cdf.timeseries AS ts ON CONCAT('FirstnameBirthyear:', dp.sensor) = ts.externalId AND ts.dataSetId = '1234567890'
Hey @Aleksander Poverud, I figured out that the trick is adding AND ts.dataSetId = '1234567890' at the end
here is the full query:
SELECT concat('FirstnameBirthyear:', dp.sensor) AS externalId, cast(time_stamp / 1000 as timestamp) AS timestamp, cast(value AS double) FROM IFSDB.values AS dp, --selecting from _cdf.timeseries means we select from the timeseries we ingested to CDF earlier. We do this to make sure all the time series we try to add data points to actually exist _cdf.timeseries AS ts WHERE concat('FirstnameBirthyear:', dp.sensor) = ts.externalId and ts.dataSetId = '1234567890'
I realized my mistake, I need to replace Matt1996-AvevaNet with my dataset ID that was recorded earlier. Problem solved :)