Functional Data Explorer - additional role expanding "Z, Supplementary"We use Functional Data Explorer (FDE) both for predicting scalars from “curves” as well as calculating DoE based factor impact on curve response. All of these additional variables, DoE factors AND Scalar response(s) AND sometimes identifier/name are added to the Z, Supplementary role in the FDE platform call. If running Functional DOE within the FDE platform all supplementary variables will be regarded as factors independent of their correct role. Therefore, Functional DoE cannot be used inside the platform. It is needed to Save Model Summaries that carry along all Z, Supplementary and then run further analysis manually in the Summaries table where we can assign Supplementary variables to their "correct" role. It would be helpful if it were possible to assign the supplementary variables’ roles into different supplement areas depending on their role in the overall analysis.
Statistics and AnalyticsUser InterfaceRe: Functional Data Explorer - additional role expanding "Z, Supplementary" - Status changed to: AcknowledgedThanks for this request. We are looking into this.
Re: Functional Data Explorer - additional role expanding "Z, Supplementary"@blum_b_1 I know it has been a few months since you posted this, we are catching up on our Wishlists, so I apologize for the delay. As I read the post, I am not quite sure what the issue is? Could provide more detail or an example data table the illustrates the problem?
Re: Functional Data Explorer - additional role expanding "Z, Supplementary"@SamGardner If running FDE, the columns assigned to "Z, Supplementary" will be carried along. After having fit the data series by use of X, Input, Y, Output and ID, Function we can run Functional DOE Analysis from the FDE's report red triangle. This FDOE will however use all variables in Z-Supplementary as factors in the Generalized Regression approach.
What I'd really want it to do could be the following:
- run FDOE GenReg with factors aql_len, core_len, core_wd, landing_zone_mat_factor, core_shape_factor1 --> factors will predict FPCs and by that will model curve shape
- run a GenReg with extracted FPCs as factors and aql_wd as scalar response or core_proposition as binomial response --> FPCs/curve shapes will model scalar/binomial response
- use p_product just as identifier/label variable with no role in any of the fitting exercises. --> no modeling involved.
Re: Functional Data Explorer - additional role expanding "Z, Supplementary"Thank you for that explanation. To clarify your request, it sounds like to me you want the following:
- Specify a set of factors to use for FDOE in Functional Data Explorer.
- Note: you can be more specific about what model to use by created a Model script in the data table, and FDE will pickup those factor and use that as the starting model in FDOE. This is described in the documentation.
- Specify other columns that you want to have carried along in the analysis to use for other purposes in FDE output tables. Those columns could include columns used a factors for FDOE.
- Specify other identifying or grouping columns to be carried along in the analysis output tables from the Functional Data Explorer platform.
For 2 and 3, it is possible to do this by joining the FDE output with the data from the original file, but it is an extra step in the analysis workflow.
Is that an accurate summary @blum_b_1 ?
Re: Functional Data Explorer - additional role expanding "Z, Supplementary" - Status changed to: Acknowledged
