Be sure to review our Idea Submission Guidelines for more information!
Submission GuidelinesHello,
After used the new "Image Recognition Tool" a few days, I think you could improve it :
> by adding the dimensional constraints in front of each of the pre-trained models,
> by adding a true tool to divide the training data correctly (in order to have an equivalent number of images for each of the labels)
> at least, allow the tool to use black & white images (I wanted to test it on the MNIST, but the tool tells me that it necessarily needs RGB images) ?
Question : do you in the future allow the user to choose between CPU or GPU usage ?
In any case, thank you again for this new tool, it is certainly perfectible, but very simple to use, and I sincerely think that it will allow a greater number of people to understand the many use cases made possible thanks to image recognition.
Thank you again
Kévin VANCAPPEL (France ;-))
Thank you again.
Kévin VANCAPPEL
We need some way (unless one exists that I am unaware of - beyond disabling all but the Container I want to run) to fire off containers in particular order. Run Container "Step1" then Run Container "Step2" and so on.
In some of our larger workflows it's sometime tedious to run a workflow in order to see some data, when adding something in the beginning of the workflow. Running und stopping it as soon as the tools gets a green border is sometimes an option.
It would be convenient to have an option in the context menu to run a workflow only until a specific tool.
In effect, only this specific tool has an output visible for inspection and only the streams necessary for this tool have been run - everything else is ignored and I'm fine to not see data for the other tools.
This would speed up the development of small parts in a larger workflow much more convenient.
Regards
Christopher
PS: Yes, I can put everything else in a container and deactivate it. But a straight forward way without turning containers on and off would be preferable in my opinion. (I think KNIME as something similar.)
Make the Container Caption Font Size Adjustable
I find it helpful to see the entire workflow at once. It would be very helpful for the container size font to be adjustable. For example, I am documenting a workflow with many containers and tools. The containers represent segments of my workflow. When I am looking at or printing the entire workflow, the container heading is too small to be read. If the font size were adjustable, it could be increased to be readable and still fit easily into the length of the container.
Thanks to zuojing80 and tcroberts for their comments on 9/10/2018.
@AdamR_AYX did a talk this year at Inspire EU about testing Alteryx Canvasses - and it seems that there is a lot we can do here to improve the product:
https://www.youtube.com/watch?v=7eN7_XQByPQ&t=1706s
One of the biggest and most impactful changes would be support for detailed unit testing for a canvas - this could work much like it does in Visual Studio:
Proposal:
In order to fully test a workflow - you need 3 things:
To do this:
This would allow:
Example:
For this canvas - there are 2 inputs; and one output.
Each test case would define:
This would make Alteryx SUPER robust and allow people to really test every canvas in an incredibly tight way!
Love the functionality to create filters on the Calgary database but it would be nice to be able to select the columns you wanted returned. There are times where you only want a couple columns but the input tool will return all columns creating a larger dataset then required. You can add a select right after the input but this is after the entire dataset has been loaded into memory. Combining the two would make the Calgary input tool behave more like a database then a standard "dumb" input source.
Please add the ability to specify indexes when creating a table with the Write Data In-DB tool.
When running Teradata SQL using the Connect In-DB tool I need to create a table on the database using the Write Data In-DB tool and do numerous updates before bringing the data to the PC. Currently there is no way to create a unique primary index (or any other index) when the Write Data In-DB tool creates a table. This causes Teradata to consume huge amounts of wasted space. Today I created a table with 160 columns and 50K rows. This consumed over 20 Gigabytes of data with 19.7 Gigabytes of wasted space. In Teradata the way to control wasted space(skew) is by properly defining the index which can't be done today.
It would be a handy feature if it were possible to choose a data type for an input tool to read the data in as. For example, if a dataset has multiple fields with different data types, it would be handy to be able to make the Input Tool read and output them all as a string, if needed. This would also make a handy tool, a sort of blanket data conversion to convert all fields to the specified type.
default file path in "File Browser" interface app would be a nice to have feature. Similar to what we have in Numeric, Text etc. interface app.
This should be a simple addition:
In the same way that many default tools automatically generate annotations when they are dropped into the workflow, or their configurations are modified, there should be a way to add custom annotations for custom made macros.
Thanks!
Both Input and Output tools should have the ability to read or write any file type from/into standard compression types (ZIP and GZIP). This would be helpful when managing large files.
It would be great if you could create default settings for the Tool Containers. As workflows become larger, I use containers a lot. But once I have 10-15 containers, I have to set all of them to have a Transparency of 1 and a margin of None. While the changes don't take long to make, it would be nice if they could be preset.
The designing interface is very slow when we design an in-db workflow.
The reason of that is that Alteryx connects everytime he needs to refresh the data. Example on Hive :
Mar 20 15:28:49.453 DEBUG 6048 HardyConnection::Connect: Default branding specific auth mech: 2 Mar 20 15:28:49.453 DEBUG 6048 HardyHiveClientFactory::CreateClient: Create HS2 client. Mar 20 15:28:49.453 DEBUG 6048 HardyHiveClientFactory::GetBackendCxnPool: Create session manager. Mar 20 15:28:49.453 DEBUG 6048 HardyHiveClientFactory::GetBackendCxnPool: Create backend connection pool. Mar 20 15:28:49.453 DEBUG 6048 HardyHiveCxnPool::GetHS2Cxn: Create HS2 connection. Mar 20 15:28:49.453 DEBUG 6048 HardyHiveCxnPool::GetCxnFactory: Create backend connection factory. Mar 20 15:28:49.453 DEBUG 6048 HardyHiveCxnFactory::CreateHS2Cxn: Create HS2 HTTP transport. Mar 20 15:28:49.453 DEBUG 6048 HardySessionManager::GetSession: Getting new session handle. Mar 20 15:28:50.399 DEBUG 6048 HardyTCLIServiceThreadSafeClient::OpenSession: TOpenSessionReq client_protocol = HIVE_CLI_SERVICE_PROTOCOL_V1
Maybe we could have an option on the IN DB Connection configuration to stay connected while designing (maybe with a limit time).
(PS : we also tried the option to Disable Auto Configure, it's clearly not he solution)
As you may know, the interrogation of Hive to get the Metadata is actually very slow on Alteryx
A first step of improvement (at least in the Visual Query Builder) has been proposed here
But the real issue for Hive is that the way Alteryx queries the Metadata : it passes "Show table" queries for all the databases. On our cluster, it means more than 400 queries that last each avout 0.5 seconds. The user has to to wait about 4 minutes.
A solution : using an API in java to ask the Hive metastore if it exists (it may be an other tab in the In database configuration). Our cluster admin has an example of a Thrift API in java that we can give you.
Result : 2 seconds for a 38700 tables in more than 500 databases !!
At present - to identify the dependencies of your workflow - you have to go to to "Advanced Settings" to find this critical capability.
(see @MattB 's great post here: https://community.alteryx.com/t5/Alteryx-Knowledge-Base/Workflow-Dependencies/ta-p/49696 )
Could we instead move this to the workflow properties on the left hand side - this would be a more logical place to keep this info.
CC: @rijuthav; @jithinmony; @HengHe; @RajK; @ydmuley; @revathi; @Deeksha; @MPistone; @Ari_Fuller; @Arianna_Fuller; @JoshKushner; @samN; @avinashbonu; @Sunder_Sriram; @Rahul_Thakur; @Rahul_Sing
It would be really nice if we could save our own custom color palette when coloring tool containers and comments.
I use colors to define the purpose of my tool containers and it would be much easier if I could select a labeled, reusable color.
I would like to see a time interface tool similar to the Date and Numeric Up Down tools. I am working on some macros where the user can select the time they would like to use a filter for the data.
Example: I want all data loaded after 5:00 PM because its late and needs to be removed.
Example 2: I want to create an app where the user can select what time range they would like to see records for (business hours, during their shift, etc)
Currently this require 2-3 numeric up downs or a Text box with directions for the user on how to format field with Error tools to prevent bad entries. It could even be UTC time.
I've seen several posts and questions concerning NULL dates. Is 09/31/2010 a valid date? I know that 02/29/206 isn't valid and that 02/00/2006 isn't either, but I really don't like finding out about these in conversion warning messages.
I might suggest a function that returns True or False on the date check and let the user configure appropriate rules to rethink the attempted date prior to committing the field to the date data type.
Cheers,
Mark
In some cases, the information about incoming columns to tools are (temporarily) forgotten, e.g. if Autoconfig is switched off, if the incoming connection is temporarily missing, or if column names are generated dynamically and the workflow has not been executed, yet.
Many tools deal with that situation well, e.g. Selection, Formula, or Summarize. In these cases, the tools tell the user that they cannot find incoming columns, but they preserve the configuration so that the user still can (at least partially) work on these tools and important information on the configuration is not lost:
Example Select Tool
Other tools behave the opposite, for example Unique or Macro Input (an for sure many other tools). If the incoming columns are currently unknown to the Designer and you click once on the symbol, the entire configuration of this tool is lost. You might try to get the configuration back by pressing undo. This, in most cases does not work. Or, even worse, you find out what happened later when it's too late for undo. In this case, you either have an old version of that workflow to look up the configuration or you have to re-develop it. In any case, this is unnecessary and time-consuming software behaviour.
Example Unique Tool
I wasn't sure whether I should report this as a bug or a feature enhancement. It is somehow in between. Two aspects tell me that this should be changed:
Please make sure that all tools preserve their configuration also if information on incoming columns is temporarily lost.
Today, there is an checkbox to "Disable All Tools that Write Output" within the Runtime settings for a workflow. Setting this option requires at least 3 clicks:
Could a keyboard shortcut be added for this? I've spoken to several users who leverage this feature and, while it is already a time saver, it seems helpful enough where a keyboard shortcut is warranted.
User | Likes Count |
---|---|
25 | |
9 | |
6 | |
6 | |
5 |