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
Hey Alteryx!
Here's one. I don't care for the tabs at the top showing each workflow. They are not in order by open date and it's hard to tell whether tab all the way on the right is the final workflow or not... They are hard to X out (especially the last one on the right that seems to have a semi hidden X). Depending on your monitor layout/resolution, you can only see like 10 visible workflows at a time in the gui. If you want to see more, you have to go into the workflow dropdown carrot which AFAIK also doesn't seem to be in any particular order esp when you open up several flows having same name. On a side note - what happened Alteryx? They USED to be in order by open date going across at the top AND in the workflow dropdown... That was useful.
So let get rid of these tabs. It's so windows95.
Instead create an interactive workflow history table (window) similar to what exists in TOAD, sql assist and other database tools (see image below). It's a table in the view but each row would reflect a workflow (formerly a tab). Top row is most recent workflow and moving down the table shows history of workflows.
You could can hover over a row to see details pop up for that workflow OR you could click each row (double click to run?) and workflow becomes visible in active canvas and overview and results window update accordingly. Maybe we could even have more than one active window per session? (sort vert/horz).
Additional benefits/functionality:
1. Slightly more screen real estate which we know you are always looking for and asked about 🙂
2. This could also be an autosave feature. When Alteryx crashes - and it does - when you reopen designer all your history will be there - no loss of productivity! Also high on your list.
3. Have an additional search function where we can search for tool names, comments, annotation, workflow/app/macro names, odbc connections, specific formulas or key words and then we see which history rows match our search criteria (rather than searching using notepad++ 'find in Files' feature in a directory)
4. It's not that much dev time since workflow history is already available in some buried log so why not show some of this metadata to the user in the gui?
Thanks,
Simon
When I open I'd workflow, I'd like the option click on "something" to see content previous created for the workflow. The content I'd use most frequently is: How was the INPUT data obtained? ALL of the Alteryx workflows I will be creating with my current employer will require some level of pre-build of the data sources. I would like to document the pre-build requirements (could be pages of text, so "comments" aren't optimal) and be able to ask for them as needed.
Optimally, this feature would accept .docx, .pdf and other standard formats. Another option would be to support a URL that would link me to the content. For my purposes, this would be even more advantageous as all of our knowledge content is on-line.
One of things I like to do is select a bunch of tools and to see the collective performance of particular part of a workflow. It would be super cool if these stats were aggregated.
I could imagine then you could run alternative solutions side by side and then just select the tools involved to compare.
Currently, Control Containers only report in case they are activated:
Info: Control Container (3): Control Container Activated.
...
Info: Control Container (3): Control Container Completed.
It would be very helpful to also get information on disabled Control Containers:
Info: Control Container (3): Control Container Deactivated.
Currently, Cache and Run Workflow cannot be used if there's a second branch of data which goes around the tool which you want to cache. This slows down intermediate runs especially in complex workflows with a lot of inter-dependencies.
For example:
If you want to cache at the formula "Do something", Alteryx will not allow this because a single temporary cache file at this tool will not be sufficient:
My suggestion is that Alteryx allows multiple cache files for one single cache event. In this case:
Proposed process:
This way, I could use Cache and Run Workflow much more often which would save me a lot of time.
When the user has no internet connection or when the Help > Source setting is set to "Offline", clicking on a custom tool's help icon will always navigate to
which is practically never desirable for custom tools.
Specifying an offline help URL
Developers should have a way to (optionally) specify an offline help URL. This configuration could behave the same as `help_url` which already supports specifying both URLs and local files. If the offline help URL is not set, it would fallback to the `help_url` configuration.
{
"tools": {
"MyCustomTool": {
"configuration": {
"help_url": "https://mywebsite.com/help/my-custom-tool",
"offline_help_url": "file://help/my-custom-tool.html"
}
}
}
}
All built-in tools have their names, descriptions, help URLs, etc. localized to the currently selected locale. Custom tools should also be able to do this.
Specify localized configuration values
Developers should have a way to specify localized values for any locales they choose to support. Similar to the deprecated .yxlang files, this can be done with supplementary `ayx_workspace.[locale].json` files that redefine a subset of configurations which support internationalization.
/* ayx_workspace.jp.json */
{
"tools": {
"MyCustomTool": {
"configuration": {
"long_name": "カスタムツール",
"description": "これは日本語です",
"help_url": "https://mywebsite.com/help/jp/my-custom-tool"
}
}
}
}
All built-in Designer tools have an associated example workflow that can be opened by right-clicking a tool and selecting Open Example. This feature is invaluable in helping users familiarize themselves with a tool and explore all of its features. Custom tools should be able to specify example workflows of their own.
It is currently possible to specify an example workflow, but
Specifying an example workflow
Developers should have a way to specify the example workflow associated with any custom tool. This would likely be done in the `ayx_workspace.json` file.
{
"tools": {
"MyCustomTool": {
"configuration": {
"example_path": "configuration\\MyCustomTool_1_0\\example.yxmd"
}
}
}
}
Note that a workflow may depend on some external files as well. It would be ideal if an `assets_path` could be defined to easily allow any arbitrary assets to be included with your tool. (This is currently possible, but must be done by manually modifying the .yxi archive; it isn't handled by the SDK). This would easily allow for example workflows to be specified as a path relative to `assets_path`, and if the workflow contains any relative paths to files within the assets folder, they can be resolved.
{
"tools": {
"MyCustomTool": {
"configuration": {
"assets_path": "configuration\\MyCustomTool_1_0\\assets",
"example_path": "example\\example.yxmd"
}
}
}
}
Localization of the example workflow would also be supported when localization of the plugin configuration is supported.
When a plugin is created, it can be labelled with any tool category, existent or non-existent. If the category does not exist, Designer will create it and give it a default icon. It would be nice if a category icon could also be specified, specifically for this case.
Specifying a tool category icon
Developers would have a way to (optionally) specify the tool category icon for their plugin. This would likely be part of `ayx_workspace.json`.
{
"name": "My Plugin",
"tool_category": "Custom Category",
"tool_category_icon_path": "configuration\\tool_category_icon.png",
"package_icon_path": "configuration\\package_icon.png",
...
}
Behavior when installing a plugin
If a plugin has a tool category icon specified, Designer should do the following when installing the plugin:
This prevents changing the icons of existing categories (including built-in categories), but would prevent subsequent plugin versions from updating their tool category icons.
Use case
The company Acme Corporation develops a plugin for interfacing with Acme-related products, and would like all its tools to be grouped in a new category called "Acme" whose icon is the Acme icon.
It would be nice to create our own custom categories to be able to organize tools. It would be similar to how the Favorites category works, but any number of additional categories can be created.
Creating a new category
The user would have a way to create a new category and enter the following information:
Adding and removing tools from custom categories
The user would have a way to add tools to or remove tools from the custom categories.
In many programming languages, errors can be thrown and caught so that the final program does not fail even if a procedure or function produces an error.
I would like to propose something similar for Alteryx, using the Control Containers. Already now, a Control Container outputs the log in the container to its output stream on the upper right corner:
Still, the workflow produces an error:
My idea is to optionally avoid the error output to the log but only report the error through the control container's output log. The error message in the output log should be "degraded" to a warning. The output log of the control container should stay unchanged. The configuration of the Control Container could look like this (addition in red):
This would help a lot if stuff produced by others is used (e.g., third party macros) which might fail while this might be normal behaviour for the workflow.
いつも大変お世話になっております。本日は、製品の動作環境についてご相談させていただきたく、ご連絡いたしました。
弊社のお客様の中には、Windowsのユーザー名やフォルダ名に日本語を使用している方々が多く存在します。
現状、こうした日本語環境でのAlteryx製品の動作保証が明確にされていないため、問題が発生した際の対応に困難を感じております。
したがって、Windowsの日本語ユーザー名、または日本語のフォルダ名を使用した環境における
Alteryx製品の動作保証について、再検討をお願いできますでしょうか?
これにより、お客様へのサポートがより円滑に行えると考えております。
お忙しいところ恐縮ですが、具体的な方針や計画についてご教示いただければ幸いです。何卒、よろしくお願い申し上げます。
All the input tools like Input Data and Dynamic Input will have a new flag "Skip on fail" that will process all the data, or none of the input data, or partial of the data requested and will return the data that could be read and do not return any error in the WFs.
If the 'Skip on fail' flag is false - the system should act like it is now.
if the 'Skip on fail' flag is true - the system should return the only the accepted or manager to read data on the default out put, and can have a second output connection for the error log, so we can parse it and do something with it, but the WFs should still run,
Some of the predictive tools put out a "Score" field when output is run through the scoring tool, and some put out a "Score_1" and/or "Score_0". Since I frequently reuse the same workflow template for different predictive model types, it would be nice if they were consistent so that I wouldn't have to crash the workflow the first time through to get the input field names correct for downstream tools (e.g., Sort). Thank you
The default variable size is a V_WSTRING of size 1073741823. If no one catches this, it uses up the memory on the server. Could the default be smaller?
Hi,
Today I pressed F1 in the Output Tool to find out what the setting Transaction Size does. It turned out that this is not documented.
It would be a great idea to make a documentation that covers the options of a tool.
Regards,
Frank
It would be very helpful to have a find and replace work in duplicated formula. especially if it is long.
Having the ability to specify columns to be "coalesced" (essentially filled in automatically when records aren't present) across all inputs of the Join Multiple tool would be very useful, and save lots of tedious setup. In fact, I avoid that tool if possible, currently, because of the extra steps necessary in maintaining key column integrity after the join.
In the screenshot below "RecordID" is present in all 4 inputs. When they are joined together, however, not every record is present in all inputs, therefore each of the 4 "RecordID" columns has nulls. We then have to go through an extra step to find the first "non-null" record ID, and put that in a new column (or in the first "RecordID" column).
This could be simply fixed by adding a menu to specify columns that should be coalesced, or auto-filled from other inputs if it's missing in the first one.
The columns in the menu could be based on the first input, and then alteryx would look for columns with the same name in all other inputs and complete the logic below, for instance:
If isnull([RecordID]) then
if isnull([Input_#2_RecordID]) then
if isnull([Input_#3_RecordID]) then
[Input_#4_RecordID]
else [Input_#3_RecordID] endif
else [Input_#2_RecordID] endif
else [RecordID] endif
Hi,
I was thinking that this might be nice addition - while Joining two inputs at the bottom there is always possibility to check if we want to include Unknown columns. Maybe we could specify if we want to join Unknown columns only from Right Input or Unknown only from Left Input. I know I would use this in my workflows.
Thanks
Alicja
Wouldn't it be great to be able to pick results from a drop down, based on the up-stream tools in the workflow? I had this situation many times, where I had to create a complex, chained app, just because the tools connected to the interface can't run before the interface tools are displayed to the end user.
For example, imagine an app, that based on what column it sees, it lets you drop one by just picking it from a drop down. It would open many development opportunities, and decrease the number of chained apps we need to build.
User | Likes Count |
---|---|
29 | |
13 | |
7 | |
7 | |
6 |