Coordinate Sorted vs Read Sorted BAM files give drastically different results

So I've noticed that when using bam files as input to pairtools parse, it matters if it is straight from the aligner, coordinate sorted, or read name sorted. This is my workflow -

pairtools parse \
    --assembly assembly_name \
    --chroms-path sizes.txt \
    --drop-sam \
    coordinate_sorted.bam | \
pairtools dedup \
    --mark-dups | \
pairtools select \
    "(pair_type=='UU') or (pair_type=='UR') or (pair_type=='RU')" | \
cooler cload pairs \
    --assembly assembly_name \
    -c1 2 \
    -p1 3 \
    -c2 4 \
    -p2 5 \
    sizes.txt:1000 \
    - \
    output.cool

If I use the coordinate sorted bam, i.e. the output of samtools sort, then I get the following matrix -

[[1 0 0 ... 0 0 0] [0 1 0 ... 0 0 0] [0 0 0 ... 0 0 0] ... [0 0 0 ... 0 0 0] [0 0 0 ... 0 1 0] [0 0 0 ... 0 0 3]]

If I use the read name sorted bam, i.e. the output of samtools sort -n, then I get the following matrix -

[[233 233 6 ... 2 0 0] [233 423 241 ... 1 0 1] [ 6 241 392 ... 2 2 1] ... [ 2 1 2 ... 642 282 19] [ 0 0 2 ... 282 448 217] [ 0 1 1 ... 19 217 406]]

I have the following questions -

1) Is this expected behavior? If so, perhaps there should be a warning.
2) I believe normally the output of aligners is read grouped, does this mean I can expect the same results from the regular bam output and the read sorted bam?

open2c / pairtools

Coordinate Sorted vs Read Sorted BAM files give drastically different results #102