This site uses different types of cookies, including analytics and functional cookies (its own and from other sites). To change your cookie settings or find out more, click here. If you continue browsing our website, you accept these cookies.
We have discussed on several occasions and in different forums, about the importance of having or providing Alteryx with order of execution control, conditional executions, design patterns and even orchestration.
I presented this idea some time ago, but someone asked me if it was posted, and since it was not, I’m putting it here so you can give some feedback on it.
The basic concept behind this idea is to allow us (users) to have:
Repetitive patterns to be reusable.
Select after and Input tool
Get not matching records from join
Tell Alteryx to execute some logic if something happens.
Any other condition
Order of execution
Need to tell Alteryx what to run first, what to run next, and so on…
Run this first
Execute this portion after previous finished
Wait until “X” finishes to execute “Y”
Putting all together
This approach involves some functionalities that are already within the product (like exploiting Filtering logic, loading & saving, caching, blocking among others), exposed within a Tool Container with enhanced attributes, like this example:
The approach is to extend Tool Container’s attributes.
This proposition uses actual functionalities we already have in Designer.
So, basically, the Tool Container gets ‘superpowers’, with the addition of some capabilities like: Accepting input data, saving the contents within the container (to create a design pattern, or very commonly used sequence of tools chained together), output data, run the contents of the tools included in the container, etc.), plus a configuration screen like:
Refers to the actual interface of the Tool Container.
Provides the ability to disable a Container (and all tools within) once it runs.
Idea based on actual behavior: When we enable or disable a Tool Container from an interface Tool.
Input and output data to the container’s logic, will allow to pickup and/or save files from a particular container, to be used in later containers or persist data as a partial result from the entire workflow’s logic (for example updating a dimensions table)
Based on actual behavior: Input & Output Data, Cache, Run Command Tools, and some macros like Prepare Attachment.
Order of Execution: Can be Absolute or Relative. In case of Absolute run, we take the containers in order, executing their contents. If Relative, we have the options to configure which container should run before and after, block until previous container finishes or wait until this container finishes prior to execute next container in list.
Based on actual behavior: Block until done, Cache, Find Replace, some interface Designer capabilities (for chained apps for example), macros’ basic behaviors.
Conditional Execution: In order to be able to conditionally execute other containers, conditions must be evaluated. In this case, the idea is to evaluate conditions within the data, interface tools or Error/Warnings occurrence.
Based on actual behavior: Filter tool, some Interface Tools, test Tool, Cache, Select.
Notes: Documentation text that will appear automatically inside the container, with options to place it on top or below the tools, or hide it.
This should end a brief introduction to the idea, but taking it a little further, it will allow even to have something like an Orchestration layout, where the users can drag and drop containers or patterns and orchestrate them in a solution, like we can do with the Visual Layout Tool or the Interactive Chart tool:
This has probably been mentioned before, but in case it hasn't....
Right now, if the dynamic input tool skips a file (which it often does!) it just appears as a warning and continues processing. Whilst this is still useful to continue processing, could it be built as an option in the tool to select a 'error if files are skipped'?
Right now it is either easy to miss this is happening, or in production / on server you may want this process to be stopped.
I often need to create a record ID that automatically increments but grouped by a specific field. I currently do it using the Multi-Row Formula tool doing [Field-1:ID]+1 because there is no group by option in the Record ID tool.
Also, sometimes I need to start at 0 but the Multi-Row Formula tool doesn't allow this so I have to use a Formula tool right after to subtract 1.
So adding a group by option to the Record ID tool would allow the user not to use the multi-row formula to do this and to start at any value wanted.
Love the new updates to the Browse tool in 2019.2! However, if you choose the option Open results in new window, which I do often so I can see my whole dataset, the search/filter/sort functionality goes away. Would be great if that new functionality also worked in the new window. Thanks!
I'm really liking the new assisted modelling capabilities released in 2020.2, but it should not error if the data contains: spatial, blob, date, datetime, or datetime types.
This is essentially telling the user to add an extra step of adding a select before the assisted modelling tool and then a join after the models. I think the tool should be able to read in and through these field types (especially dates) and just not use them in any of the modelling.
An even better enhancement would be to transform date as part of the assisted modelling into something usable for the modelling (season, month, day of week, etc.)
Can we have a tool to optimize another tool's configuration based on an output target? For example optimize the fuzzy logic setup to find the optimal tool configuration that yields the best matching score for a given data set.
During my trial of assisted modelling, I've enjoyed how well guided the process is, however, I've come across one area for improvement that would help those (including myself) overcome any hurdles when getting started.
When I ran my first model, I was presented with an error stating that certain fields had classes in the validation dataset that were not present in the Training dataset.
Upon investigation (and the Alteryx Community!) I discovered that this was due to a step in the One Hot Encoding tool.
Basically, the Default setting is for all fields to be set to error under the step for dealing with values not present in the training dataset, but there is an option to ignore these scenarios.
Add an additional step to Assisted Modelling that gives the user the option to Ignore / Error as they see fit.
If this were to be implemented then it would remove the only barrier I could find in Assisted Modelling.
Hope this is useful and happy to provide further context / details if needed.
Assisted modeling is a great idea but right now it's a bit unflexible.
IMHO the greatest strength is the semi-automated transform tool, which would extremely helpful on its own.
What would be great is the possibility of using the features without having to go through the assisted modeling wizard but as single tools with minimal configuration, so that it could be used as an automated system for quickly choosing variables.
This way we could have a pretty much perfect rapid prototyping tool for machine learning tasks, leaving more freedom in modeling and enabling less skilled analysts in easily finding on which variables they should focus.