forum.smartpls.com

hi Jan-Michael,

thanks for clarifying this. Getting results later has never occured to me but then I was maybe not patient enough. In that case an option to unhide the window would be helpful. Just a thought, I got used to the current behaviour ;)

BR! Jan

on my win 8.1 system when i export the model as PNG or SVG the file is saved without a file definition. e. g. if I type "mymodel" the file is saved as "mymodel" instead of "mymodel.png". Same applise for .svg.

Hi smartPLS-Team, I'd suggest that you put a "all calculations are finished" message in bold at the end of all other calculation messages. At the moment you only have a "[successful]" in the tab-name which is easily overlooked. Also I'd think you should be clearer what pressing &...

Thank you for clarifying this!

Thank's for your answer. But that is not exactly what I wanted to know. My situation: Number of cases: 1222. Search Depth in settings: 1000 Used Search Depth in Report: 40. So in this case I don't need to re-run the algorithm with Search Depth 1222 because thix maximum was not reached anyway and the...

Just to make sure, I get PLS-POS Settings right: Before running the algorithm I can set "Maximum Iterations" and "Search Depth". After running the algorithm I find the actual values in the report POS-Settings. In case "Iterations used" and "Search Depth used" ...

read footnote 6 of that paper: Other distance-based methods, which are in earlier stages of development and currently not available as software packages, include fuzzy PLS path modeling for latent class detection (FPLS-LCD; Palumbo et al. 2008) and partial least squares genetic algorithm segmentatio...

With FIMIX you explore and search for homogeonous groups, with PLS-MGA you verify that groups are actually different. If you have your model and want to verify that there are no hidden sub-groups that differ significiantly from each other you can use FIMIX (and PLS-POS) to let algorithmns search for...

I suggest you get your hands on a copy of "a primer on partial least squares" by Hair et al. as they describe it in general. Apart from that I don't think a forum is the right place to answer such general questions, although I would be delighted if somebody would create a video to help oth...

I suggest you ask one of the authors via https://www.researchgate.net/

What do you mean by run? Making the Multi group analysis or running them in standard PLS Algorithm but separately for each group?

Hi Choukri,

it is quite simple: You select the column as column 0 that hold's the gender variable and leave the others unchanged. As Prefix in your example take "Gender_". then press ok. Now you see a news tab "data groups" wherer your data set is split by Gender.


BR, Jan

You need to edit the smartpls.ini file in the smartPLS folder: set the xmx value to a larger value (this is the maximum amount of RAM smartPLS is allowed to use - should be less than your physical ram): my value is: -Xmx4096m -Xms512m (xms is the minimum value, don't think you really need to change ...

there is a free implementation of PLS in R:
http://cran.r-project.org/web/packages/semPLS/

a lot less comfortable but from reading the manual it seem's that it can also be used for moderator effects

Hi Elise, having a semi-translated model and no specific questions is maybe not the best start to get feedback. Also making use of abbreviations doesn't foster replies (what is e-loyalty in?). As a start reading "A Primer on Partial Least Squares Srtuctural Equation Modeling (PLS-SEM)" hel...