Hi
I'll answer each part after the question.
ANONYMOUS wrote:
> Good afternoon Professor.
>
> Here are some questions about examples.
> 1. In the project description you provide a sample document sample2.tsv. Is this document only for testing with best_predictor?
I posted it primarily as input to best_predictor (particular in relation to the correlation calculation). However, it also serves to show what the output to the cleaning stage looks like.
> 2. During the actual detection, whether a tsv file is provided, processed by cantril_data_cleaning, and then given to best_predictor for testing. Or will different tsv be provided for these two scripts for detection.
Different tsv files, derived from the data cleaning stage, will be used to test the analysis program.
> 3. Is it possible to check whether best_predictor is correct by directly running ./best_predictor sample2.tsv? That is, if my own output is different from the example you provided, it means there is something wrong with my script.
You don't need the output from my script for sample2.tsv. Just take the correlation values for each predictor metric versus Cantril in the Sample1.tsv Excel spreadsheet. (That is, for a given predictor, the output correlation value will be the average of the two values for those countries with valid data.)
Cheers
MichaelW
👨🎨