i don't know how to use lsqcurvefit in this case

1 次查看（过去 30 天）

Meddour Aissam riad 2020-2-20

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/506610-i-don-t-know-how-to-use-lsqcurvefit-in-this-case

Hi everyone,

so these is my problem:

let suppose i have this equation Y=P*(a*x+b)

i have a data vector for x let's say x=[1 2 3 4 5 6 7 8 9] , and the initial value of P let'sz say P=5, and i need to find the the value of a and b that minimise the squared error between my function and another vector (Not an Y vector) with the function LSQCURVEFIT.

this is how my algorithm should work, my LSQCURVEFIT should give me a value for 'a' and 'b' and with those value i have to calculate the ydata for all the value of 'x'

i'll take the last value of Y and i should stock it in a new vector let's name it 'ANA'. the last value of my ydata will replace the value of my 'P' and then i should do the same calucation for 20 times for exemple (so we will use a for loop)

so in final i have to compare the value of thisvector (20 number) with an experminetal data that have also 20 number

how should i procceed , if you need more explanation i'm here