Zeroes reported as 0.0e+00

[Gustavo Velásquez]

The new number formatting system is causing some zeroes to be reported as 0.0e+00.

> tab(mtcars, gear, vs, am)
 
    gear │       vs │       am │    Freq.  Percent     Cum. 
─────────┼──────────┼──────────┼────────────────────────────
       3 │  0.0e+00 │  0.0e+00 │       12    37.50    37.50 
       3 │        1 │  0.0e+00 │        3     9.38    46.88 
---------┼----------┼----------┼----------------------------
       4 │  0.0e+00 │        1 │        2     6.25    53.12 
       4 │        1 │  0.0e+00 │        4    12.50    65.62 
       4 │        1 │        1 │        6    18.75    84.38 
---------┼----------┼----------┼----------------------------
       5 │  0.0e+00 │        1 │        4    12.50    96.88 
       5 │        1 │        1 │        1     3.12   100.00 

> sum_up(mtcars, gear, vs, am)
 
Variable │      Obs  Missing     Mean   StdDev      Min      Max 
─────────┼───────────────────────────────────────────────────────
    am   │       32        0  0.40625  0.49899  0.0e+00        1 
    gear │       32        0   3.6875   0.7378        3        5 
    vs   │       32        0   0.4375  0.50402  0.0e+00        1 

[Matthieu Gomez]

thanks. solved by 8115649

[Matthieu Gomez]

Just to expand a little bit on the current formatting system, the function format_fixedwidth_number(x, w) is supposed to mimic di %w.0g x in stata.