- ago
If (by accident) the historical data provider "Random" is checked in Data Manager->Historical Providers it is possible to request any symbol from this provider (returned as synthetic random data).
I consider this potentially dangerous!
I think the "Random" provider should return data only for the symbols that were entered when a new random data source is created, where I tend to use names like "RND1", "RND2" and s on.
I had the situation, where I have a DataSet (a large set of ETFs) marked as "Data Provider Linked" with WealthData and still the backtest used data from the Random provider for symbol "SJB".
This of course leads to very strange results and some headaches.
Workaround: make sure the Random provider is disabled!
Here a step by step instructions to reproduce:
* Enable WealthData provider and Q-Data Provider
* Enable the Random Provider
* Open a Chart (File->New Chart)
* Enter symbol "SJB"
There appears a dropdown that offers the symbol "SJB" from three providers:
It is not possible to reliable select a specific provider.
Sometimes WL gets "stuck" with the Random provider and display this random data.
The same happens in a backtest: Sometimes WL chooses the Random provider.
Please fix this, it took me weeks....
I consider this potentially dangerous!
I think the "Random" provider should return data only for the symbols that were entered when a new random data source is created, where I tend to use names like "RND1", "RND2" and s on.
I had the situation, where I have a DataSet (a large set of ETFs) marked as "Data Provider Linked" with WealthData and still the backtest used data from the Random provider for symbol "SJB".
This of course leads to very strange results and some headaches.
Workaround: make sure the Random provider is disabled!
Here a step by step instructions to reproduce:
* Enable WealthData provider and Q-Data Provider
* Enable the Random Provider
* Open a Chart (File->New Chart)
* Enter symbol "SJB"
There appears a dropdown that offers the symbol "SJB" from three providers:
It is not possible to reliable select a specific provider.
Sometimes WL gets "stuck" with the Random provider and display this random data.
The same happens in a backtest: Sometimes WL chooses the Random provider.
Please fix this, it took me weeks....
Rename
- ago
#1
My 2 cents - operation by design is not a bug. Accidentally selecting something you don't want is a user error.
Random Provider is a Historical Provider that returns random data. It's easy to avoid - never check it. Even when it's not checked you can use it in DataSets for testing by creating your linked Random Provider DataSet and selecting it for the backtest.
If the chart is really showing data from Random instead of WealthData as in the image - that's certainly a bug, but that does not seem to be the case! The data in that chart is from WealthData.
Not being able to select the symbol in the symbol selector is a misunderstanding of how it works.
I'll let Glitch take your next response.
Random Provider is a Historical Provider that returns random data. It's easy to avoid - never check it. Even when it's not checked you can use it in DataSets for testing by creating your linked Random Provider DataSet and selecting it for the backtest.
If the chart is really showing data from Random instead of WealthData as in the image - that's certainly a bug, but that does not seem to be the case! The data in that chart is from WealthData.
Not being able to select the symbol in the symbol selector is a misunderstanding of how it works.
I'll let Glitch take your next response.
I agree with Cone, it’s working as designed.
Your Response
Post
Edit Post
Login is required