Rebasing a dataset

[dean.hong]

Hello, I would like to rebase a graph to a certain data point. I ran a regression model and I got a beta coefficient (which in this case is 5.2). I would like to rebase my graph with my beta coefficient that 0 becomes my beta coefficient (5.2). I would like to replicate this process. Thanks in advance.

[EViews Gareth]

You'll have to explain in more detail.

[dean.hong]

Each plot is a component to growth. I want to see how each component contributes to the growth level. This model is based on regression, which I also ran using EViews, and each component is multiplied by its corresponding beta to better reflect their weights. However, because beta constant coefficient is too large(5.2), my component to growth model had a huge discrepancy compared to the real growth. To solve the issue, I allocated my beta coefficient equally to each component. Now my components to growth model has the right number, but it does not truly reflect each component's contribution to growth as I have allocated my beta coefficient.

To put more easily,
I ran a regression model: growth = b0 +b1*c1 + b2*c2+b3*c3+b4*c4+b5*c5
where as c=component, b=beta.
My original model did not reflect the growth level as it did not contain (and I did not want to contain) b0. But now that I know I need to include b0, rather than allocating b0 into other components, I want to rebase my graph with my beta constant coefficient. Below is an attachment for your reference. Thanks.

[EViews Gareth]

You haven't explained what you mean by rebase.

You also haven't explained what the graph is - how it was created etc...

[dean.hong]

If I was not clear, the attached image file should explain what I mean by rebase.

I allocated beta constant (bhat(1)) into the components, which I am trying to fix. The first image file attached is the result of the allocation. The allocated beta makes it difficult to see how each component is contributing to the growth level.

caihat is the blue line. Notice how I did not add the beta here and this is making difference between the blue line in the upper graph and the composite value of the lower graph (they should be same). To remove the difference, I want to add back the beta, but not the way I did before. I want to make the BASE (which is 0 now, to 5.2, which is my beta in this case). The attached image file will explain in more detail. It might be simple if I simply want to change those numbers by hand, but I am running this model with tens of country in regular basis, so I want to know how I will be able to change the base into a datapoint (Beta) I want. Thank you.

[EViews Gareth]

I guess I'm just not seeing the issue. Why can't you just add 5.2 to the numbers?

[dean.hong]

Because if I just add the number (5.2) without rebasing as you suggested, I get the graph I am attaching now. Because the beta is too high, it is difficult to look how each component is contributing to the growth. The goal of this graph is to see how each component is contributing to the "exact" growth level. I decided to divide 5.2 by 6 (the number of components) and allocate equally to each component, but I ended up getting the graph that I posted first, and it does not reflect the growth level, as each component data contain beta/n. I would love to know if there is a way to rebase 0 to the beta coefficient that I have if I can just change the label of the tick mark. Thanks.

[EViews Gareth]

Then I'm afraid my understanding of your issue is not good enough to help. I still have no idea what you're doing.

[dean.hong]

Thanks for the response and I appreciate your help. Let me put this way. Is there any way I can re-label tick marks? Even if you cannot understand the logic behind it and why I'm doing it, I hope you understood what I am trying to do. I just want another label and that's it. Please see the attached image. I tried my best to be clear as possible. If there is any function, code or anything that can do anything like that, please let me know. Very much appreciate it.

[EViews Gareth]

No you cannot, but that's because it is generally trivial to change the numbers in the graph, rather than change the labels.

[dean.hong]

Then I guess I'll have to stick with what I have. Thank you for the answer.