Alteryx Designer Desktop Ideas

I often have to cache my workflow at certain points to do further development/analysis since the run time is so long. I can't express how frustrating it is when i need to edit a formula tool that's like 2 tools behind the cache but the whole entire cache is lost when i have to edit it. why can't the cache be kept up until the tool that was edited??

Business problem

Alteryx's block until done tool is a terrific way to sort which records go to which stream first.  However, during the last couple of years my team has experienced several times where there are gaps in functionality.  For example, we output data into a relational database (SQL Server) with foreign key constraints.  In this manner, one output tool must be completed (the dimension tables) before the others (fact tables) to properly load data.  Block until done doesn't always work in these cases, and our team has frequently relied on using chained apps or other clunky workarounds.  This creates unnecessary complexity for a simple problem.

A potential solution

Allow an optional setting in the output tool that waits for another tool in the same workflow to complete first.

Performance profiles work at the tool level. When I want to evaluate the performance of a group of tools, I have to click on them one at a time, log the performance, and calculate manually. I want to be able to click on a container full of tools or lasso some myself and view the granular and subtotaled performance profile.

Our company requires a password change every 3 months, which wreaks havoc bc i have to manually change each workflows input data connection configuration every time this happens to put in my latest credentials. Not the most efficient process for all the automated workflows I have on the gallery.

Instead, it would be great if alteryx could load in usernames and passwords via a field from the workflow. This way, I could just update the one excel file with two fields for username and password and every workflow referencing it with an input data connection would automatically pass on the new credentials every 3 months, not disturbing any of my autoscheduled workflows.

When building out Alteryx workflows there may be a need to read in different ranges within the same Excel spreadsheet. For example bringing in a table from Sheet1, but also isolating a table name in a particular cell (in my example cell C8).

When turning this into an analytic app, with a file browse is to add an action tool with the default value of "Update Input Data Tool".

However when specifying this option within the analytic app interface, you are only allowed to chose one option of the following:

i) Select a sheet

ii) Select a sheet and specify a range

iii) a named range or

iv) a list of sheet names.

The problem is in the example above I need a sheet and a range, but I want to avoid adding two file browse interface tools as it shouldn't be needed. If the user selects (i) then it loses the reference to cell C8, but I would imagine a lot of users as they get started with apps don't realise this is what will happen.

There is however a way to solve this currently and it requires overwriting the default behaviour and configuring the second action tool (the one that updates the file for C8), to update value with a formula, where you assume the user would select sheet name and then use this formula:

replace([#1],"$`","$C8:C8`")

However I would argue that this has a lot of technical debt, plus if the user needs to modify where the header is, for example to D8 they need to change the input file and the action tool so it works as a workflow and an analytic app.

Solution

Like how the configuration options for the input file, such as which row to input data from or whether first row contains data is maintained, modify the behaviour of the default option in the action tool to maintain references to ranges.

ARIMA, ETS, TS Forecast, TS Covariate Forecast are difficult to differentiate, more distinct icons would improve differentiation on the canvas. 

All the other file types have different coloured icons and it works well to differentiate them in a directory. Its impossible to differentiate workflows and database icons currently. 

Ther should be a low-cost read-only licence for the Alteryx desktop designer.

Read-only means: I can use the designer normally, but I cannot save workflows. So, for example, I can adjust the path in the Input Tool and follow exactly how the data goes through the workflow, I can even edit the workflow, but I can't save my changes.

Reason:
Alteryx is a very high-priced product. An own licence is only worthwhile if I use Alteryx very often.

But many users do not want to design workflows themselves at the beginning, but perhaps just run specific workflows. These users are usually not provided with a licence by the management, but only those who have many use cases with Alteryx. In practice, however, these are usually those who are already Citizen Developers anyway.

With a read-only licence, you could provide these users with workflows, but at the same time give them the opportunity to take their first steps with Alteryx. Many will want to be able to save their changes and apply for a licence upgrade.
Note: with Alteryx Gallery Server  workflows can be made available for execution. But on the one hand there is no interactivity and on the other hand it is very time-consuming to make workflows available in a user-friendly way in the Gallery Server.

Hi,

The basic table report tool could be improved by including a title option, such as a checkbox that allows you to add a title above the table.

When you want to add a title to a table, you must create a report text specifically for that purpose. The more significant improvement would be for situations in which you want to create tables grouped by a specific field. It would be nice to have the option to add a title per group as well, this would make it easier to add all of them to a report in a more dynamic way without having to do workarounds.

My testing has shown that when a datetime field is input from a Snowflake table, the Input tool will convert the data to reflect the datetime as local time to the machine the workflow is run on. 

For example, this is a data set from a direct query to Snowflake...

This is what is coming out of the Input tool with the machine set to Pacific time...

However when setting the machine to Central time, the results from the Input tool are...

This obviously can wreck havoc with inconsistencies of subsequent reporting off of this data. 

Since Snowflake carries the UTC offset as part of the data, it would be nice to have the ability to disable this "assumption" by Alteryx that the results should convert the datetime to the machine datetime. This way the data could pass through with the datetime values that are held in the database.

To add the capability to hard rename the columns in all modes 

Would like to be able to reference the UserID of the person running the workflow within the workflow itself, usually for authentication purposes.

For example, we use the Publish to Tableau Server tool. The main developer will embed their password in the tool and then publish it to Gallery. We are wanting to authenticate if the person running the workflow on Gallery can actually publish to Tableau Server before publishing, not just the person who published the workflow in the first place. 

Another example is that we are needing to upload data to our data lake through APIs and need to pass in user information of who is publishing to that package through Alteryx, and check that they can indeed publish there. 

Basically, we need to have logic within the workflow that is referencing who is running the workflow. 

We understand that this would most likely only be supported when workflows are run on Gallery, as there isn't a UserID tied to someone when running on a local machine. 

When you have an Alteryx workflow open, Alteryx seems to by default try to keep you up to date on what might be happening with your data when it runs through your workflow. So if you for example add a misconfigured tool (a filter not connected to an input) and click somewhere on the canvas it'll presumably try to compile the code and then figure out that the new tool is misconfigured and it'll tell you why. A major thing it does seems to be that it tries to figure out if macros included in the workflow have changed and to take such changes into account so that it can notify you if there's a problem somewhere e.g. with the macro's output schema or whatever. I know it's doing this kind of thing because the moment I add a macro to the workflow I'll have to spend a 15-20 second 'tax' every time I touch the workflow canvas, a formula, when I click on a join, etc. Sometimes it's 30 seconds, sometimes you get lucky and it'll only be 5 seconds. This delay is by now from my perspective considered a fixed cost of adding a macro to a workflow. I'm assuming similar processes also take place in the context of other dependencies (main one probably being queries inside input tools) and that they may also cause problems for similar reasons.

I'm assuming part of the reason for the long delays is that the macro repository where we usually save macros in my organisation is saved in a server location which is close to the Alteryx server executing the in-production workflows/macros, but not close to me when I'm developing in my office. Yes, I could save the macros I develop elsewhere (locally) and then only save them in the repository when they're 'complete' (...we all know exactly when that's the case; we're never in doubt about that, right? ...and you'll still have problems if you need to modify a workflow which includes macros later, even if you're not touching the macro itself). I'm actually doing that in some contexts where the above user experience has been frustrating enough to justify such a step, and I'm always trying to find ways to just outright kill Alteryx' live connection to the macro (e.g. by caching the output) if it's not critical. But these things are not solutions, they're poor workarounds some of which are adding complexity and the potential for errors as a result of a problem which really shouldn't be a problem.

It would be desirable to have the option to pause these kinds of 'background processes'/'semi-live compiling'/'whatever', make Alteryx do this kind of thing less frequently, add an 'only update meta-data when running' option, or some fourth option of a similar nature. Debugger-mode is implicitly always on, why not give the option of turning that off if the user figures s/he can handle that? Give me the error when I try to run the workflow, don't try to have the software figure out if the code will run with an error every time I even touch it - this is not always helpful, it's in some contexts causing a huge waste of developer time.

I'm testing out the new Data Connection Manager (DCM) and think it needs 1 enhancement based on the way we'd use DCM. Whenever Alteryx opens, it should sync the Data Sources/Credentials from the Server. This is critically important when sharing these with more than 1 user when a password needs updated.

For example, DesignerA updates the password in the source system, and then updates the password in their Designer DCM settings. In order for this new password to get synced to DesignerB, 2 things have to manually happen: DesignerA would need to sync the new password to the server, and then DesignerB would need to sync the new password from the server. I can live with the first part where DesignerA needs to sync to the server. It's just part of the password update process. The second step though seems perilous. DesignerB should get the new password without having to do anything; as things currently stand, DesignerB will have the old password until they manually intervene and sync the password. Imagine a scenario where it's just DesignerB, but hundreds of people who would all have to sync their credentials. 

I also think this idea makes sense in light of the way Data Connections currently work (pre 21.4) where a similar sync happens automatically every time alteryx opens.  

It would be nice to have an option to distribute tools with the fixed default space of three in between each tool. Now it just distributes based on the available space, which can be inconsistent throughout the workflow.

The idea is to have a Run option, where the workflow runs everything up to the selected tool (Like the Cache functionality does).

You select the tool, hit Run Up and the workflows executes everything "before" the selected tool.

That'll make developing much easier, specially when dealing with big workflows and constant changing data.

We're not too happy with the Gallery Data Connections not being available for the IN-DB data input tool but that will hopefully be a feature to be looked at in future product improvements; Let us know if there are reasons not having this feature already.

Thank you.

I'm Left handed and use my mouse in my left hand.  As such I use the following right-hand keyboard operations in the programs like Excel etc; CTRL-INS (Copy), Shift-INS (Paste), and Shift-DEL (Cut).

In Alteryx Designer these don't work for me in the Workflow window (but do in say the meta Info window) - is there a setting so I can enable them for the Workflow Window, or can the feature be added.

A "Filter" that would work like a "Formula" - where you can add multiple criteria in one space, and for each criteria, you would get an output anchor.  I use Alteryx to manage master data from several factories - each needing to have separate workstreams.  Stacking Filter criteria functions, but it would be much cleaner to have it managed within a single tool.