Questions regarding Source finding
Hello, I am Eslam and I am working as a data scientist at IDIA (http://www.idia.ac.za/) with a computer science background (dont know much about astronomy), currently I am busy making a tutorial that is based on the solution for this data challenge (DC1), and I have the following questions:
1- The solution shows that we perform source fining twice. One the training and the other are on the full image. Now, Is there any duplication between the two data frames, I am worried that there will be data leakage (some of the training samples are on the testing samples). I have tried to verify this but did not come to a conclusion.
2- Is it necessary to perform source findings twice [training, full]? I keep thinking that it will be easier just to do it on the full image. And we can just split the data frame from the full image into training and testing.
That is all for now, Thanks in advance