Hallo.
I have a DOE design with response data as a table, created starting from an Excel table. Here my DOE experiences:
I can create a contour plot easily. The response column - as well as all factors X - are appearing in the lists for the corresponding variables and can be used for the plot. I practically get the same plot with OriginLab. If I want to obtain an analysis diagram or a contour analysis, factors are not appearing in the lists for the axes to use them for the plot. Is it normal that this is not possible? Why?
I also created plots for the dependency of the response variable with respect to every factor. They are looking like those automatically produced in OriginLab and I produced a "Wirkungsflächendiagramm". - a surface plot -
and I got this result with JMP
However, with OriginLab I got this surface plot:
Compare the pictures: although axis parameters are the same, the surfaces do not appear to be the same.
I didn’t find the following features: I need residuals plots e.g. the dependency of the residuals on the run/ row number. I wanted a presentation of the dependency of the standardised Effect on the factors or factor interaction combinations. However, most important are presentations of the dependencies of the response on factor interactions / factor product combinations. How can I obtain all such dependencies? Is there no automatic production of such important DOE diagrams in JMP?
Another problem is, that a design can perhaps be created in JMP, analysed etc., but how to do this with a given design which has not the same order as the design created by JMP? I would need to add all response variables manually to each line which is very prone to error. However, the production of the same design with other software packages and with JMP (same algorithm with the same random seed) is still a difficult issue. Please help me to use JMP efficiently. How can I start with a given Excel design and perform a complete data analysis of a screening design with results given by an Excel file? I tried to start with the option of the menu Excel analyse diagram but it didn’t help. I also started with design diagnosis bit this didn’t*t help either. The option analyse in the menu seems not really to be helpful for DOE although run aways might be found.
Is there no easier way with JMP PRO than to do all this manually? Is this the same for the normal Version JMP? Do I need to write scripts to get automatically all results as in RStudio? Then I also could use RStudio. The OriginLab DOE App does not contain Definitive screening designs yet and also some nonlinearity features are missing, but all features mentioned above are produced automatically without any line of coding in some seconds. And manual creation of diagrams can become important work if you have 5 or 7 factors. Even the coding of the corresponding scripts could get more complicasted.
Of course I found "evaluate design", but why are there "anticipated target values" which are coming from nowhere? The response values present in the design are not used and - as far as I see - cannnot be added somewhere. If the "anticipated target values" could be replaced or if the real response variable would be directly used, the DOE faculties would be much better. Anyway, why are thes target values? First of all, the observed responses need to be used.
Sincerely
Thommy7571
