0

I have the following equation:

t ~ a1 * log(X/X0) + a2 * log((Xmax - X)/(Xmax - X0))

I used the package minpack.lm with the data frame of experimental data below, to perform a non-linear regression(nlsLM), and obtained the values of X0, Xmax, a1 and a2.

t   X
0   2.42254
5   2.64426
10  3.34135
20  5.16827
30  6.34615
40  7.45192
55  7.25962
70  8.43750

    a1         a2         X0       Xmax 
  3.132196 -34.323302   2.060305   9.547698

Now I would like to use the same model to predict the values of X, for t ranging from 0 to 70. When using predict(), I get only t values.

Since it is not possible to isolate X from the equation, how can we estimate its values using R?

NelsonGon
  • 13,015
  • 7
  • 27
  • 57
Victor Sartor
  • 11
  • 3
  • Have you checked for conflicts with the function `predict`? This is loaded in `stats`, too. [Check for conflicts using `conflicts()`](https://stackoverflow.com/questions/6354697/list-of-masked-functions-in-r) – Jessica Burnett Apr 06 '19 at 17:01
  • I believe that the `predict ()` function returns the values of `t` due to the organization of the model (`t = "..."`). I considered isolating `X`, but as I mentioned above, this does not seem possible. – Victor Sartor Apr 06 '19 at 17:17

0 Answers0