I use the following codes to model batch effect for differential analysis from different batches. 

ta <- dba(sampleSheet = "diff.csv")
ta <- dba.count(ta, summits=75)
ta <- dba.contrast(ta, categories=DBA_CONDITION, minMembers = 2, 
                   block=DBA_REPLICATE)
ta <- dba.analyze(ta)
ta.DB <- dba.report(ta, contrast=1, method=DBA_DESEQ2_BLOCK, bCounts=TRUE)

When I plot the PCA and heatmap after it, it shows the error:

 "Error in pv.getPlotData(DBA, attributes = attributes, contrast = contrast,  :  Only one site to plot -- need at least 2!".

Any advice? Thanks!

Ultimately, the differential binding plots are failing because there isn't enough differential binding being exhibited by your data.

As the error states, there is only a single site identified as being significantly differentially bound using the default thresholds. DiffBind will not plot heatmaps or PCAs using only a single site.

You can try lowering the FDR threshold to see more sites (by setting the th parameter in the plotting functions), The report will show how many sites will be identified as differentially bound using alternative thresholds.

Cheers-

Rory