Thank you so much @benjjneb ! Now I understand why there is a difference when change the order of my sampleIDs. And I agree that this difference is not a big problem.
I have sent you an e-mail attached with the imported sequencing data on Monday. But I am not sure whether I can send e-mails with large files (>500M) to Gmail in China. Maybe I have been blocked... Sorry for the delay and it seems you do not need the data anymore.
And I have to say sorry to @ebolyen . I was supposed to install the latest R and dada2 quickly and finish that analysis. However, I have been struggling with updating R and installing dada2 via BiocManager for four days. I feel so frustrated. But I promise once I
solve this I will change the order of sampleID and increase the error model reads number to affirm Ben Callahan's conclusion.