In my conversation service, I'm using Watson conversation service. In dialog nodes, one branch has sub-nodes. When I train the bolt, it didn't go to the sub-node I defined, so I picked the one I defined, but I got "Unable to change the intent. Error: Unique Violation". What does this error mean? How to fix it?
Thanks for your help!
Emily
I have seen such an error when you try to add a question to an intent that is a duplicate to an all ready defined question within the same intent. It also happens when you try and define a duplicate to an Entity.
Double check your questions. Otherwise I think we will need more info or examples of the code etc.
Related
I'm trying to add new record to my DB, but I got this error:
This is my schema.rb:
And this is my method create:
And this is my form to add a new record:
The param is sending correctly, but when I click on Upload button, nothing is happen, and if I try to add the record from the rails console, I got the error of the first image. I'm a new dev of RoR, someone can help me please?, Thanks!!
Without looking at the related models and their associations nothing is clear, because, it seems like your model associations are causing this error.
Model associations depict the relations between models (if its one-one, one-many and such) known as inheritance, very important concept in any object oriented programming. this is the way we implement it in rails
Go through this documentation.
It might help you get better understanding of rails associations.
Could someone point me to documentation that shows how to create a record in a model with one to many relation via client side script.
Thanks in advance!
There are two ways of creating a new record.
1. Via a client-side script
You can do this by calling a createItem function from a widget. This link might help with that.
2. Via a server-side script
You can do this by directly creating a new record, assigning it to the relation of a model and then saving that model. This link might help with this method. By the way, it's easier to do this in a server-side script, as the guide itself says.
Since you just requested the link to official documentation, I've limited my answer to links only. Please feel free to ask if you need clarifications!
I have a UBE that is failing after making a call to a jdeObjectCall. The API returns a 1, indicating a warning, which indicates to the report to stop processing.
The image below shows the jdeObjectCall in question that is returning a 1.
I'd like to determine why jdeObjectCall is returning a warning, so that I can investigate a fix.
Any help on this is greatly appreciated!
Cheers, Cam.
I was able to determine the answer to this question with some help from the forum jdelist.com. The topic in which I discovered my answer can be found here: https://www.jdelist.com/vb4/showthread.php/53758-E1920-Debugging-jdeObjectCall
The way I found my solution is as follows. Apologies for the high-level steps. There is some assumed knowledge here. If you are interested in a more in-depth explanation of one of these items please comment and I will provide it.
Enable logging in the jde.ini file.
Debug the locally.
Examine the output of the log file using a tool like JDETrace.
You can do a search for your business function within JDETrace. In my case, I searched for EditDate.
The line in the log will let you know the name of the busines function object.
Checkout the object in object management workbench.
Find the line where the EditDate function exists.
Place a breakpoint on this line and continue debugging.
I am trying to use the Watson REST APIs to add examples to an Intent for Watson. Before I call the Create Endpoint I call the Get Example endpoint with the intent and example.
When I call the Get Example endpoint with the word "fine" it returns a 404. Then when I try to Create an example it returns a 400 response
{"error":"Unique Violation: The value \"fine\" already exists"}
This is happening because we already have an example "Fine" (notice the first letter is capital).
How can I prevent this? Are there best practices to store examples in all lower case? Or should I just catch the 400 exception and look at the error.
Violation error means it didn’t update. So you can certainly check for that to take action. Although I personally recommend looking for the related item first to avoid the error.
Coding convention recommendations. These formats are used to easily recognize what is referenced in code and if an identifier is missed.
For example is this below an intent, entity or context variable?
accountingPayBillCode
Intents
All caps, spaces as underscores.
#ACCOUNTING_PAY_BILL
The examples (questions) should be entered untouched as how you received them. Do not attempt to fix spelling / grammar errors.
Example:
I need to pay my bill. Can yuo help me?
Entities
CamelCase with first word capitalized. The value should be all lowercase, and avoid multiple words (but must be meaningful).
#AccountDetail:code
The reason to avoid multiple words as the value is that you can end up with something like this.
#AccountDetail:(part number)
It makes it more prone to a mistake.
Synonyms should also be stored all in lower case.
Context variables.
Always reference using the $ prefix. Use camelCase with first character lowercase.
$accountCode
In my dialog flow I have an node triggered by an #app_question intent that answers with a question about a person's location - "Where's your location?", then, I've created a child node that is triggered by an entity #app_location inside of which I have #app_location:americas, #app_location:other.
Watson is answering correctly in the Try it out chat, when I say I'm in america, it outputs the correct answer, but "Irrelevant" appears in the node box. Should I worry about that, is that standard or I am doing something wrong?
ps: I was using Jump to, but received an error. If I'm supposed to use Jump To, I can copy and paste the error here.
Below are screenshots:
If anyone can help, thanks a lot!
Conversation will always try to determine what the intent is, even if you are not interested in it at that time. The try it out is just showing you what it tried to figure out what the intent was. Not that it would do anything with it.
As you are looking for an entity and not an intent it should be fine.