[GRASS-dev] Incomplete raster history in r.mapcalc

GRASS GRASS developer
1

hi,

by chance I observed a detail:

GRASS 6.4.2svn (utm49n):~ > r.mapcalc "L5120060_06020060925_ndwi = 1.0
* (L5120060_06020060925_toar.4 - L5120060_06020060925_toar.5) /
(L5120060_06020060925_toar.4 + L5120060_06020060925_toar.5)"
100%

-> 1.0

GRASS 6.4.2svn (utm49n):~ > r.info -h L5120060_06020060925_ndwi
Data Source:

Data Description:
   generated by r.mapcalc
Comments:
   1 * (L5120060_06020060925_toar.4 - L5120060_06020060925_toar.5) /
   (L5120060_06020060925_toar.4 + L5120060_06020060925_toar.5)

-> 1

? I wonder where this get's eaten...

Markus

2

Markus Neteler wrote:

by chance I observed a detail:

GRASS 6.4.2svn (utm49n):~ > r.mapcalc "L5120060_06020060925_ndwi = 1.0
* (L5120060_06020060925_toar.4 - L5120060_06020060925_toar.5) /
(L5120060_06020060925_toar.4 + L5120060_06020060925_toar.5)"
100%

-> 1.0

GRASS 6.4.2svn (utm49n):~ > r.info -h L5120060_06020060925_ndwi
Data Source:

Data Description:
   generated by r.mapcalc
Comments:
   1 * (L5120060_06020060925_toar.4 - L5120060_06020060925_toar.5) /
   (L5120060_06020060925_toar.4 + L5120060_06020060925_toar.5)

-> 1

? I wonder where this get's eaten...

  static char *format_constant(const expression * e)
  {
      char buff[64];
  
      if (e->res_type == CELL_TYPE)
    sprintf(buff, "%d", e->data.con.ival);
      else
    sprintf(buff, "%.8g", e->data.con.fval);
  
      return strdup(buff);
  }

"%g" omits trailing zeros, and omits the decimal point if nothing
follows it. Using "%#.8g" would retain the decimal point and the
trailing zeros (G_trim_decimal() could be used to remove those).

--
Glynn Clements <glynn@gclements.plus.com>