Using splines to find the Area Under the Curve

Using splines to find the Area Under the Curve - In these two worksheet the Area Under the Curve is found for a biomass batch reactor and a pulse - Mensajes

#1 Publicado: 15/5/2023 20:34:26
Javier Ochoa

Javier Ochoa

3 me gusta en 26 mensajes.

Grupo: User

Hi,

These two examples solve the problem calculations for the total biomass formed in a batch bioreactor and the pattern flow of a reactor with a pulse of some tracer. The splines method and the Simpson integration formula found in the forum let me solve both problems.


Kind regards,


splines_integral_data.sm (30,28 KiB) descargado 1667 vez(es).

splines_integral_data1.sm (31,52 KiB) descargado 1642 vez(es).
A 1 usuarios les gustó este mensaje
Oscar Campo 26/5/2023 7:10:00
#2 Publicado: 22/5/2023 15:02:07
CBG

CBG

78 me gusta en 312 mensajes.

Grupo: User

Hello

Fitting Javier's data to the Gauss function by the Levenberg-Marquardt Algorithm

Fit_Javier_LMA.png

Javier_fit_CBG.sm (37,14 KiB) descargado 1632 vez(es).


Best Regards

Carlos
A 1 usuarios les gustó este mensaje
Javier Ochoa 22/5/2023 15:20:00
#3 Publicado: 22/5/2023 15:56:44
Javier Ochoa

Javier Ochoa

3 me gusta en 26 mensajes.

Grupo: User

Hi Jean and Carlos,
These worksheets help me to better understand the problem which I tried to solve.

Thanks in advance,

#4 Publicado: 23/5/2023 16:29:20
Javier Ochoa

Javier Ochoa

3 me gusta en 26 mensajes.

Grupo: User

Hi Jean and Carlos,

I am curious about your thoughts for the result I found using the al_nlqsolve routine from ALGLIB.

May be I was lucky to find a similar solution.

Kind regards,


LevenbergMarquardt_solver.sm (27,68 KiB) descargado 1682 vez(es).
#5 Publicado: 24/5/2023 0:30:27
CBG

CBG

78 me gusta en 312 mensajes.

Grupo: User

Wrote

Hi Jean and Carlos,

I am curious about your thoughts for the result I found using the al_nlqsolve routine from ALGLIB.

May be I was lucky to find a similar solution.

Kind regards,


LevenbergMarquardt_solver.sm (27,68 KiB) descargado 1682 vez(es).



Hi Javier
As you can see in the graph, the green line is that of your solution,
and it is not correct, since it does not fit the data.

javier_fit2.png

It is always important to draw the curve

Javier fit.sm (24,31 KiB) descargado 1618 vez(es).

Best Regards

Carlos

A 1 usuarios les gustó este mensaje
Javier Ochoa 24/5/2023 9:31:00
#6 Publicado: 24/5/2023 3:41:54
Alvaro

Alvaro

1.139 me gusta en 1.805 mensajes.

Grupo: User

Hola Javier. Your Levenberg Marquardt algorithm works pretty well, but you choose a bad guess starting point, with a very high tolerance. BTW, I made some few modifications and use also other target function.

LevenbergMarquardt_solver.sm (32,85 KiB) descargado 1662 vez(es).
LevenbergMarquardt_solver.pdf (233,35 KiB) descargado 1000 vez(es).

Best regards.
Alvaro.
A 2 usuarios les gustó este mensaje
Javier Ochoa 24/5/2023 9:31:00, Вячеслав Мезенцев 24/5/2023 4:21:00
#7 Publicado: 24/5/2023 15:52:25
Alvaro

Alvaro

1.139 me gusta en 1.805 mensajes.

Grupo: User

Wrote

Hi,

These two examples solve the problem calculations for the total biomass formed in a batch bioreactor and the pattern flow of a reactor with a pulse of some tracer. The splines method and the Simpson integration formula found in the forum let me solve both problems.


Kind regards,


splines_integral_data.sm (30,28 KiB) descargado 1667 vez(es).

splines_integral_data1.sm (31,52 KiB) descargado 1642 vez(es).



Hola Javier. You can also use this code for integrate and differentiate tabulated data.

Click to enlarge

splines_integral_data.sm (85,66 KiB) descargado 1659 vez(es).

splines_integral_data.pdf (180,67 KiB) descargado 933 vez(es).

Best regards.
Alvaro.
A 1 usuarios les gustó este mensaje
Javier Ochoa 29/5/2023 9:25:00
#8 Publicado: 25/5/2023 13:16:26
Jean Giraud

Jean Giraud

983 me gusta en 6.866 mensajes.

Grupo: User

.. refreshed/augmented.
Cheers ... Jean.

Inst_Frechet GAUSSIAN.sm (92,39 KiB) descargado 1715 vez(es).
A 1 usuarios les gustó este mensaje
Javier Ochoa 29/5/2023 9:25:00
  • Nuevos mensajes
  • Sin mensajes nuevos