no valid observation

[morag]

hi.

for some reason i keep on getting this warning when i try to estimate the equations (as seen at the file).
i tries ti estimate this equations:
verbo=c(2)*jerusalem+c(3)*haifa+c(4)*northcenter+c(5)*southcenter+c(6)*south+c(7)*TLV+c(8)*grapefruityellow+c(9)*grapefruitred+c(10)*lemon+c(11)*pom+c(12)*pomred+c(13)*klem+c(14)*klemichal+c(15)*klemtapuzina+c(16)*tapuz+c(40)*automn+c(41)*spring+ c(42)*winter+c(44)*rain +c(45)*mintemp+c(46)*maxtemp+c(47)*protest-c(48)*lnprice+c(49)*lnsjg_2 @ pricesito _1 _2 _3 _4 _5 _6 q_average jerusalem haifa northcenter southcenter south TLV grapefruitred grapefruityellow lemon pom pomred klem klemichal klemtapuzina tapuz automn spring winter rain mintemp maxtemp protest

lnprice=c(50)*jerusalem+c(51)*haifa+c(52)*northcenter+c(53)*southcenter+c(54)*south+c(56)*TLV+c(57)*grapefruityellow+c(58)*grapefruitred+c(59)*lemon+c(60)*pom+c(61)*pomred+c(62)*klem+c(63)*klemichal+c(64)*klemtapuzina+c(65)*tapuz+c(66)*automn+c(67)*spring+ c(68)*winter+c(73)*water+c(74)*foe+c(75)*protest+c(76)*katiff+c(77)*pricesito+c(78)*num2011+c(79)*num2012+c(80)*num2013+c(81)*num2014+c(82)*rainfruit+c(83)*fruitmintemp+c(84)*maxtempfruit+c(85)*salary-log((c(48)*(1-@logit(c(49))*sj_g2-(1-@logit(c(49)))*sjt2)/c(48)*(1-@logit(c(49))*sj_g2-(1-@logit(c(49))))*sjt2)+1-@logit(c(49))) @ q_average jerusalem haifa northcenter southcenter south TLV grapefruitred grapefruityellow lemon pom pomred klem klemichal klemtapuzina tapuz automn spring winter water foe protest katiff pricesito num2011 num2012 num2013 num2014 rainfruit fruitmintemp maxtempfruit salary

or any other equation. now it keeps on giving me this note. i tried to upload the version. to put a start values. and i just dont understand what ealse can i do.

thank you a lot!

hadar only2.wf1

(6.13 MiB) Downloaded 254 times

[EViews Gareth]

Almost certainly due to the @logit(c(49)) term. You need that to be a non-NA.

[EViews Glenn]

And more generally, your coefficients C cannot be NAs.

[morag]

the term @logit(c(49)) is because the model really means something only when the parameter c(49) is from 0 to 1. unless its doesnt means nothind and for some reason its not happening. so it cant be NA. is there any other way to do so? (by the way...)

what coefficent c?

and , now the error said that 'log of non positive number' (i pretty sure its about the -log((c(48)*(1-@logit(c(49))*sj_g2-(1-@logit((49)))*sjt2)/c(48)*(1-@logit(c(49))*sj_g2-(1-@logit(c(49))))*sjt2)+1-@logit(c(49))) term), and the sj_g2 and sjt2 have no non positive numbers at all. what can it be?

thanks a lot for the answer

[EViews Gareth]

Open up the C vector in your workfile and look at the entries. They have to be valid.