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
I think this is fairly self-explanatory and is a very common use case.
Would like to have a tool that can natively support DateTimeAdd functionality where you can define working days of the week for inclusion and holidays for exclusion.
Must be able to handle both adding and subtracting.
I find the concept of Batch and/or Iterative macros, when done specifically for the simple purpose of iteration, to be a fair bit of overhead. If we could extract the fundamental qualities of a loop and get that into an "Iteration Tool," it could become a well-used tool from the pallette.
Implementation Ideas:
An example problem this could solve is from: http://community.alteryx.com/t5/Data-Preparation-Blending/Looping-and-dynamically-changing-output/m-.... As discussed therein, the concept of "row dependent iteration" makes this difficult to solve with standard tools.
If the input data set from that example were sent into the proposed Iteration Tool... it would automatically loop over the dataset rows; and three expressions could be supplied in the Tool configuration to solve the problem:
VarE: IF [i] > 1 THEN VarF[i-1] + VarG[i-1] ELSE VarE ENDIF VarF: VarA + VarB VarG: VarC + VarD
For implementation purposes, this would be logically equivalent to:
VarE[i]: IF [i] > 1 THEN VarF[i-1] + VarG[i-1] ELSE VarE[i] ENDIF VarF[i]: VarA[i] + VarB[i] VarG[i]: VarC[i] + VarD[i]
(so, basically, the i'th row is assumed unless otherwise provided in the expression syntax).
I hope this isn't too outlandish - I've tried to think through how this could be accomplished (1) as a tool that is not too fiendishly difficult for Alteryx to implement and (2) which would also be easy for us, the end users, to utilize. Thanks!
It will be great if we have a round down or round up function just like excel in Alteryx. I know we can use the left function after converting the number to string, but it becomes difficult when the decimal value is huge.
I have many aggregation in the same workflow which product one or two different column each, total could be xx different columns. Wouldn't it be nice to have a multiple append tool that could take connections from many other tools? At the moment I would have to use many separate append fields tools and then one transpose, or multiple transposes and union. Either way you have to attach something to each tool.
In a short workflow, this might not be necessary as the information related to each tool is spelled out in the progress windows. However, in a complicated and lengthy workflow, tracing such msg can be a tedious task. In addition, using a tool with multiple outputs and only one output is selected while the residual outputs may be used to validate the result in the selected output; for example, joint tool where left or right output should be zero, a visual queue could be a quick way to alert operator on any potential problem. Certainly, a browse tool can be added but in a big workflow, couple with a large data set, it might be a drain to the system resource. What if there is a tool that would activate a visual alert, like a light bulb, based on a preset condition to tell user that something is wrong and perhaps additional work needs to be done to either remedy or to account for the residual data. As in the case of a joint where 100% match is desired, any unmatched row would require an update to the reference list which maybe an additional adhoc process outside the current process. Certainly, an additional steps can be added to first explore the possibility of unmatched data and to update the reference list accordingly. The workflow would in hold until 100% match is achieve. This would require additional system resource in order to hold; especially with large set of data and lengthy workflow. If the unmatched situation rarely occurs, just a lightweight visual queue that 'pop' while allow the process either to break or to go through might be a sensible solution. Just a thought.
It would be handy to have a tool that mirrored another specified tool in the workflow. I frequently use the same regex or filter configurations in multiple parts of my workflow. If I could just update one of those tools and have every other one where I want the same logic applied to automatically update, that would make things easier.
Users would just have to enter the tool ID of the tool they want mirrored.
It would be great if it was possible to output the top most influential features in producing the score for each individual entity/row when using the predictive and machine learning tools.
Similar to the way they work in DataRobot. Details here and here.
This would enable some simple interpretation of how a model came to an individual prediction and the most important features in that particular row/case.
I'd love to see a Haskell Tool similar to the already existing Python and R-Tool.
Haskell is a great way to handle data without side-effects. I know I am probably pretty much alone with this suggestion, but I wanted to share it anyway.
I also know Python can do basically everything Haskell can do, but it isn't as beautiful and effective as it is in Haskell. Haskell brings many logic operators and great syntactic "sugar" out of the box. Things like pattern matching with input parameters is just amazing in Haskell and would require way more code in Python.
Under the join ribbon, could there be a new tool added called Minus?
This would behave similar to how SQL Minus works. I think this tool could closely resemble the setup found in the union tool and provide the following functions:
When doing validations in a workflow, this would provide a useful dq measure to use as a sanity check to ensure that for example loan numbers haven't been dropped from one point of the workflow to the next.
We should have a simple mathematical functions under Formula Tool - Math functions to calculate specific Percentile data point. Or it can be a separate tool to derive various other statistical information.
The Python equivalent is Numpy.Percentile() OR quantile().
I have explored Alteryx app, help documentation and community posts without success.
I have a workflow that accepts 4 different sets of inputs.
It also has four different outputs. However, I can easily output to separate files in four containers and just close the containers that don't pertain to the input.
On the input side, I keep having to delete the connections and connect the desired input.
I would like to connect multiple files to one tool and have the tool only work if only one of the inputs it's live. Through the use of containers you would then choose which input was live.
Hi Team,
The data set has 2 columns with missing data,
Amount
Age
If the user wants to impute the
missing values in Amount column with '$0" and
missing values in Age column with "mode",
it would be nice to have a single imputation tool to perform this.
Please let me know your thoughts.
Thanks