Dave,
Per the GRASS4.1 Programmer's Manual (a recommended reference!!):
"Each grid cell is stored in the file as one to four 8-bit bytes of data."
"When the data values in the raster file require more than one byte, they are
stored in *big-endian* format, which is to say as a base 256 number with the
most significant digit first."
"Negative values are stored as a signed quantity, i.e., with the highest bit set
to 1:"
To summarize, GRASS supports integer values from -2E+32 to 2E+32. Beware of the
handling of floating point numbers within GRASS programs!
--
Malcolm D. Williamson - Research Assistant E-mail: malcolm@cast.uark.edu
Center for Advanced Spatial Technologies Telephone: (501) 575-6159
Ozark Rm. 12 Fax: (501) 575-3846
University of Arkansas
Fayetteville, AR 72701
I have Landsat TM data, Band 5 and 7. I refer to the r.mpacalc
tutorial manual that came with the 4.1(beta?) release. Following the
tutorial I use r.mapcalc to scale the two bands. The tutorial states
that the raster image is stored as an integer file. What is the data
format of the resulting new raster image ? Is it signed or unsigned 8
bit, 16 bit, 32 bit integers?EG: mapcalc> img1 = ((2.95*tmb.5) + (4.2*tmb.7)) - float(30)
I am concerned because somewhere I thought I read that GRASS stores
raster images as unsigned 8 bit words (0-255), and it's quite possible
that the equation above could generate negative numbers which would be
then probably clipped to 0.If this is the case is there an alternative data type I can select?
dave
bever@erim.org