Sanger imputation assigns some empty SNP IDs, Is this normal?

Question

[rshoobs]

Hi all! I recently performed some imputation using the sanger imnputation service. While I was parsing the data I noticed that some of the SNP IDs it outputs are filled with a period '.' rather than an actual ID.

for reference the output looks somewhat like this

##header info
##header info
#CHROM  POS     ID      REF     ALT     QUAL    FILTER  INFO    FORMAT  SAMPLE_ID1 ...
10      235413  rs2448371       G       A       .       PASS    RefPanelAF=0.665456;AN=154;AC=95;INFO=0.995195  GT:AD 
10      235386  .       G       A       .       PASS    RefPanelAF=0.000302663;AN=154;AC=0;INFO=1       GT:ADS:DS:GP ...
10      235434  rs549315358     T       A       .       PASS    RefPanelAF=0.000403551;AN=154;AC=0;INFO=1       GT:AD ...
10      235475  rs559476434     T       C       .       PASS    RefPanelAF=0.000100888;AN=154;AC=0;INFO=1       GT:AD ...
10      235494  rs528633330     T       A       .       PASS    RefPanelAF=0.000100888;AN=154;AC=0;INFO=1       GT:AD ...
10      235511  .       T       C       .       PASS    RefPanelAF=0.000100888;AN=154;AC=0;INFO=1       GT:ADS:DS:GP ...
...

I just want to highlight that some of the IDs are unfilled. To fix I was just planning on assigning any empty ID with one based on CHR:POS. However I am wondering why sanger left these IDs as empty and if they signify any problems with the data/outputs. I am planing to run MatrixEQTL on this data downstream if that is relevant.