14 factors resultater in a large design matrix, and you might want to conduct a fractional factoring design. That being said, you can find free software (small python scripts) that will automatically generate the design matrix based on the number of factors.
can you tell me what is script for generating full factorial design? I Have 68 factors, 4 runs and 2 levels, orders are randomised by the results itself.
Thanks, much appreciated. Yes, this also works for more than two levels and three factors. A primary example is the central composite design. Remember to subscribe 😊
hey, here you manually entered the high/low levels for each factor but in cases where you are dealing with more than three factors (i.e. I have 14 factors with two levels), how can one assign high/lows for each run? i.e. while still ensuring orthogonality?
Thanks! Remember to subscribe 👍 I will look into 3 factor/3 lever designs, in the meantime you might want to refer to "Engineering Statistics" by Montgomery.
yes, the number of treatment combinations in a full factorial design is p^k where p=number of levels, k=number of factors. If you have n runs of each combination (i.e. n-1 replications) the total number of runs is n*p^k
Thank you so much for your simple explanation I’ve been struggling with this for a week now
Fantastic!!!
Thanks! Remember to subscribe for more 😊
14 factors resultater in a large design matrix, and you might want to conduct a fractional factoring design. That being said, you can find free software (small python scripts) that will automatically generate the design matrix based on the number of factors.
can you tell me what is script for generating full factorial design? I Have 68 factors, 4 runs and 2 levels, orders are randomised by the results itself.
Thank you so much for this! I wonder if it can be applied the same for 3 factors with five levels.
Thanks, much appreciated. Yes, this also works for more than two levels and three factors. A primary example is the central composite design. Remember to subscribe 😊
hey, here you manually entered the high/low levels for each factor but in cases where you are dealing with more than three factors (i.e. I have 14 factors with two levels), how can one assign high/lows for each run? i.e. while still ensuring orthogonality?
Best video ever
Thanks Daniel, glad you found it useful. Remember to subscribe 😊
Excelent and simple= easy to understand.
Would you have an example for 3 factor 3 level fractional and full factorial?
Thanks! Remember to subscribe 👍 I will look into 3 factor/3 lever designs, in the meantime you might want to refer to "Engineering Statistics" by Montgomery.
@@firmacomedy Hi, I did subscribed and am looking forward for the 3^3 fractional a and full factorial example. Thank you
which method are you using for the calculations. is this the factorial plan 2^k with k the number of factors ?
If I have 6 factors; so I need to write 64 lines with the two corresponding levels?
yes, the number of treatment combinations in a full factorial design is p^k where p=number of levels, k=number of factors. If you have n runs of each combination (i.e. n-1 replications) the total number of runs is n*p^k
How do l optimise the model say exact temperature or pH or concentration to give the maximum yield?
My plan is to use solver in excel. If thats tge case for x1, x2 etc do l use -1 and 1s, or actual figures for my factors?
@@teelit7655 I have the same question, can you help if you found it out?
What if we have 5 levels for each factor? How do I arrange the (-2,-1,0,1,2)?
yes that's one way of doing it. Check "Central Composite Design" - that might be relevant to you
Which book do you refer to in this video?
Engineering Statistics, Montgomery et al.
How would you determine low and high levels
Can you elaborate on that?
I am getting a NUM! error in my p value for some reason
Thankss
Thank you
You're welcome
You have an excel file about Design of Experimpents (DoE) for Injection Molding.
Please give me that file.
Thank you very much for your help.