Avgidea Function
AFX is capable of executing any user-created Python program (function) against files stored in Avgidea Storage. Since all function execution is completed within ADP, there is no need to download or upload files to or from the client terminal. The schedule feature can also be used to execute functions at desired intervals.
By utilizing functions, data preprocessing and prediction using AI models can be performed on the storage during data analysis. Below is an example of a function implementation.
Character encoding conversion
Replacement/deletion of character strings
Extraction of a specific range of strings
Text data classification and etc.
The following four types of functions can be created and executed from the browser.
Line : The text file is read one line at a time and passed to the function as an argument
File : The file is passed to the function as an object
Directory : Storage Files under the data source are passed to the function as objects.
Standalone : Operates as an independent function, no dependency on input data
DataFrame : The CSV file is passed and received as a Pandas DataFrame.
In addition to operation from the console, it is possible to execute and stop functions via the API. In an OEM environment, at least one cluster must be configured before a function is executed.