Member-only story
Last night I was working on a dataset that did not have very well-defined feature names, and this anomaly caused the program I was working on to go into error. The solution to the problem was to use Tensorflow’s AdvancedArguments function, but I was unable to find any sample code to help me to correctly use it.
I looked Tensorflow’s advancedArguments function up on the Tensorflow documentation, but the below screenshot was all of the information that was available on this function:-
I searched the internet for some sample code on how to use Tensorflow’s AdvancedArgument to no avail. There was no sample code that I could find. I thought about posting a question on the Stackoverflow website, but I was not quite sure how I could word my question without getting a rude response.
I did not want to give up, so after taking a break from my quest to use this function, I decided that I would try to formulate the code myself.
Specifically, I needed to set fail_on_non_keras_compatible_feature_name = False, as the default for this parameter is set to True. I decided to define the variable, argument, and place the function and…
