Hi, I have panel data made of 99 cross-sectional units and 7 time series. When I do the fixed effects model in gretl, there are only 158 observations, 56 cross-sectional units and 4 time series. My model has one dependent variable and 5 independent variables. There are some values missing and I have told this shoudn't be a problem while doing the linear regression. But gretl removed all of the units where there is only one variable missing. Even when I have data for all other 5 variables, gretl removed it from model completely because on is missing. This reduced my dataset significantly. Could you please advise how to fix this?