[GRASS-dev] strange r.stats output -- bug?

[Kirk Wythers]

Can anyone explain the behavior of r.stats when applied to a raster that has been aggregated using r.resamp.stats with the "mode" option? It seems to me that the output should look like the example below from r.stats on the original resolution map (250 meter resolution) in the fist example. However, this is not the case. When applied to the aggregated map (1000 meter resolution), r.stats reports binned cell values and counts. 

GRASS 6.4.1 (northcentralus_albersequalarea):~ > r.resamp.stats -w input=mnwimifnfftgk7wt175v3_250m output=mnwimifnfftgk7wt175v3_1000m method=mode --verbose 100%

This all appears to work properly. If I export the raster to MATLAB and simply view the cell values they are all whole number like 500, 600, 700, etc....

However, when I run r.stats on each map, I get:

GRASS 6.4.1 (northcentralus_albersequalarea):~ > r.stats -cl input=mnwimifnfftgk7wt175v3_250m 100%
0  893025
100  18419
120  34024
170  125
260  18
380  527
400  4269
500  36610
600  163
700  12479
800  51090
900  53455
990  5
999  527
* no data 874

GRASS 6.4.1 (northcentralus_albersequalarea):~ > r.stats -cl input=mnwimifnfftgk7wt175v3_1000m
100%
0-3.917647 from  to  929109
97.941176-101.858824 from  to  11467
117.529412-121.447059 from  to  32643
168.458824-172.376471 from  to  1
258.564706-262.482353 from  to  1
376.094118-380.011765 from  to  30
399.6-403.517647 from  to  383
497.541176-501.458824 from  to  24231
599.4-603.317647 from  to  12
697.341176-701.258824 from  to  2184
799.2-803.117647 from  to  50355
897.141176-901.058824 from  to  55190
995.082353-999 from  to  4

The output to r.stats from the 250 meter resolution map look fine... but doesn't make sense to me that the 1000 meter resolution map would have it's count and cell values binned into groups like 0-3.9, and 97.9-101.9, when I was using the "method=mode" flag in r.resamp.stats???

What am I missing here?

Thanks in advance