Fixed-effects (within) regression Number of obs = 111
Group variable: country Number of groups = 12

R-sq:within= 0.3958 Obs per group: min = 7
between = 0.4302 avg = 9.3
overall = 0.1801 max = 12

F(9,11) = 7.73
corr(u_i, Xb)= -0.9494 Prob > F = 0.0012

(Std. Err. adjusted for 12 clusters in country)
------------------------------------------------------------------------------
| Robust
lnlng | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
lngdp | 1.667123 1.291329 1.29 0.223 -1.175072 4.509318
lngdpC |-7.242335 13.72856 -0.53 0.608 -37.4587 22.97403
lncp | 28.03549 13.88506 2.02 0.069 -2.525317 58.5963
lnpn | 3.228645 1.179608 2.74 0.019 .6323461 5.824945
lnpri |-7.798725 3.72612 -2.09 0.060 -15.99986 .4024096
lnper | -19.5632 12.61759 -1.55 0.149 -47.33433 8.207925
lndis | 7.009837 3.184696 2.20 0.050 .0003697 14.0193
lnCNG | 1.556943 .4986949 3.12 0.010 .4593232 2.654563
stab | .2999535 .1596339 1.88 0.087 -.0513982 .6513053
_cons |-125.4026 120.7818 -1.04 0.321 -391.2416 140.4364
-------------+----------------------------------------------------------------
sigma_u |6.4897628
sigma_e |1.5543253
rho |.94574967 (fraction of variance due to u_i)
-------------------------------------------------------------------------
数据是这样的，用的固定效应模型，但距离变量前的系数为正，看不明白了，距离数据我是用的航运数据乘的每一年的国际油价，不知道是不是这个的问题，不知道怎么加权才是对的。
另外，存在很多零贸易值，请问如何解决呢？
毕业论文写不出来，求帮助