hi,
how can start values (param) with the interval with this cases:
1/c(1)=[0.1; 0.6], inside interval (take values betwwen 0.1 to 0.6)
2/c(1)=]0.1; 0.6[, outside interval (not from 0.1 to 0.6)
3/c(1)=]0.1; 0.6] all values limited at 0.6
4/c(1)=[0.1; 0.6[ from 0.1 and all outhers values
5/c(1)‹0.2
6/c(1)‹=0.2
7/c(1)›0.5
8/c(1)›=0.5
9/c(1)=0.2 just take one value
i seen than i can include a line that begins with param:
param c(1) 0.2 c(2) 0.3.....etc
can you help me
starting values with interval and condition
Moderators: EViews Gareth, EViews Moderator
-
EViews Gareth
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13604
- Joined: Tue Sep 16, 2008 5:38 pm
Re: starting values with interval and condition
Not really sure what you want here.
You want to generate random numbers between those values and use those random numbers as starting values?
You want to generate random numbers between those values and use those random numbers as starting values?
Re: starting values with interval and condition
i have this model
y c x1 x2 x3
1st case:
c(1)=0.2 take just 0.2
c(2) and c(3) and c(4) between [0.2; 0.6]
2nd case
c(1)...c(4) just take value outside ]0.2; 0.6[
3rd case:
c(1)....c(4)‹=0.6
can i make that the coef take value or take the value between interval.
you understood me.
y c x1 x2 x3
1st case:
c(1)=0.2 take just 0.2
c(2) and c(3) and c(4) between [0.2; 0.6]
2nd case
c(1)...c(4) just take value outside ]0.2; 0.6[
3rd case:
c(1)....c(4)‹=0.6
can i make that the coef take value or take the value between interval.
you understood me.
-
EViews Gareth
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13604
- Joined: Tue Sep 16, 2008 5:38 pm
Re: starting values with interval and condition
ok, so you're not talking about starting values at all? You just want to impose restrictions on the final estimated values?
Re: starting values with interval and condition
i can restrict this to use wald test after estimation the model, but how can starting values in this case.
1/c(1)=[0.1; 0.6], c(1) take value just in this interval
2/c(1)=[0.1; 0.6], c(1) take value out side this interval
3/c(1)=]0.1; 0.6] all values limited at 0.6
4/c(1)=[0.1; 0.6[ from 0.1 and all outhers values
5/c(1)‹0.2
6/c(1)‹=0.2
7/c(1)›0.5
8/c(1)›=0.5
9/c(1)=0.2 just take one value
for example for the first case, c(1) should be starting value from 0.1 to 0.6, the second case c(1) should start by value out side [0.1; 0.6].
1/c(1)=[0.1; 0.6], c(1) take value just in this interval
2/c(1)=[0.1; 0.6], c(1) take value out side this interval
3/c(1)=]0.1; 0.6] all values limited at 0.6
4/c(1)=[0.1; 0.6[ from 0.1 and all outhers values
5/c(1)‹0.2
6/c(1)‹=0.2
7/c(1)›0.5
8/c(1)›=0.5
9/c(1)=0.2 just take one value
for example for the first case, c(1) should be starting value from 0.1 to 0.6, the second case c(1) should start by value out side [0.1; 0.6].
-
startz
- Non-normality and collinearity are NOT problems!
- Posts: 3798
- Joined: Wed Sep 17, 2008 2:25 pm
Re: starting values with interval and condition
param c(1) 0.3i can restrict this to use wald test after estimation the model, but how can starting values in this case.
1/c(1)=[0.1; 0.6], c(1) take value just in this interval
2/c(1)=[0.1; 0.6], c(1) take value out side this interval
3/c(1)=]0.1; 0.6] all values limited at 0.6
4/c(1)=[0.1; 0.6[ from 0.1 and all outhers values
5/c(1)‹0.2
6/c(1)‹=0.2
7/c(1)›0.5
8/c(1)›=0.5
9/c(1)=0.2 just take one value
for example for the first case, c(1) should be starting value from 0.1 to 0.6, the second case c(1) should start by value out side [0.1; 0.6].
param c(1) -0.3
That does what you said, but I doubt it's what you want.
Re: starting values with interval and condition
i mean that i can put limits on estimated coefficient values in my linear regression.
-
EViews Gareth
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13604
- Joined: Tue Sep 16, 2008 5:38 pm
Re: starting values with interval and condition
So, again, you are not talking about starting values at all?
http://forums.eviews.com/viewtopic.php?f=4&t=48&p=99
http://forums.eviews.com/viewtopic.php?f=4&t=48&p=99
Return to “General Information and Tips and Tricks”
Who is online
Users browsing this forum: No registered users and 1 guest
