Release 0 3 1

.gitignore

@@ -1,8 +1,7 @@
 .Rproj.user/
 .Rhistory
 .RData
-delete this working file.R
-engage_2.csv
 .Rproj.user
 _*
-.*
+.DS*
+.R*

.travis.yml

@@ -0,0 +1,12 @@
+language: r
+
+os:
+  - linux
+  - osx
+
+cache: packages
+
+# safelist
+branches:
+  only:
+  - develop

DESCRIPTION

@@ -1,10 +1,10 @@
 Package: seminr
 Type: Package
 Title: Natural, Friendly, Domain-Specific Language for Building PLS Structural Equation Models
-Version: 0.3.0
-Date: 2017-10-12
+Version: 0.3.1
+Date: 2018-01-03
 Authors@R: c(person("Soumya", "Ray",
-              email = "soumya.ray@gmail.com", role = c("aut")),
+              email = "soumya.ray@gmail.com", role = c("aut", "ths")),
              person("Nicholas", "Danks",
               email = "nicholasdanks@hotmail.com", role = c("aut","cre")))
 Description: A powerful, easy to write and easy to modify syntax for

R/evaluation.R

@@ -9,8 +9,8 @@ evaluate_model <- function(seminr_model) {
 
 ## Reliability -------------------------
 # RhoC and AVE
-# Dillon-Goldstein's Rho as per: Dillon, W. R, and M. Goldstein. 1987. „Multivariate Analysis: Methods
-# and Applications.“ Biometrical Journal 29 (6): 750–756.
+# Dillon-Goldstein's Rho as per: Dillon, W. R, and M. Goldstein. 1987. Multivariate Analysis: Methods
+# and Applications. Biometrical Journal 29 (6).
 # Average Variance Extracted as per:  Fornell, C. and D. F. Larcker (February 1981). Evaluating
 # structural equation models with unobservable variables and measurement error, Journal of Marketing Research, 18, pp. 39-5
 rhoC_AVE <- function(seminr_model){
@@ -102,7 +102,7 @@ cross_loadings <- function(seminr_model) {
 }
 
 # HTMT as per Henseler, J., Ringle, C. M., & Sarstedt, M. (2014). A new criterion for assessing discriminant validity in
-# variance-based structural equation modeling. Journal of the Academy of Marketing Science, 43(1), 115–135.
+# variance-based structural equation modeling. Journal of the Academy of Marketing Science, 43(1), 115-135.
 # https://doi.org/10.1007/s11747-014-0403-8
 HTMT <- function(seminr_model) {
   HTMT <- matrix(, nrow=length(seminr_model$ltVariables), ncol=length(seminr_model$ltVariables),

R/library.R

@@ -99,7 +99,7 @@ calculate_loadings <- function(weights_matrix,fscores, normData) {
 # Function to adjust for the interaction
 # Adjustment of the SD of the interaction term as per Henseler, J., & Chin, W. W. (2010),
 # A comparison of approaches for the analysis of interaction effects between latent variables
-# using partial least squares path modeling. Structural Equation Modeling, 17(1), 82–109. https://doi.org/10.1080/10705510903439003
+# using partial least squares path modeling. Structural Equation Modeling, 17(1), 82-109. https://doi.org/10.1080/10705510903439003
 adjust_interaction <- function(ltVariables, mmMatrix, outer_loadings, fscores, obsData){
   for(latent in ltVariables) {
     adjustment <- 0

R/summary.R

@@ -8,7 +8,6 @@ summary.seminr_model <- function(object, na.print=".", digits=3, ...) {
 
   model_summary <- list(iterations=iterations,
                         paths=path_reports,
-                        metrics=metrics,
                         loadings=object$outer_loadings,
                         cross_loadings=metrics$Validity$`Cross-Loadings`,
                         weights=object$outer_weights,

README.Rmd

@@ -16,7 +16,18 @@ knitr::opts_chunk$set(
 
 # SEMinR
 
-The `seminr` package provides a natural syntax for researchers to describe PLS structural equation models.
+SEMinR brings many advancements to creating and estimating structural equation models (SEM) using Partial Least Squares Path Modeling (PLS-PM):
+
+* A _natural_ feeling, _domain-specific_ language to build and estimate structural equation models in R
+* Uses _variance-based PLS estimation_ to model both _composite_ and _common-factor_ constructs
+* _High-level functions_ to quickly specify interactions and complicated structural models
+
+SEMinR follows the latest best-practices in methodological literature:
+
+* Automatically _adjusts PLS estimates to ensure consistency (PLSc)_ wherever common factors are involved
+* Ajusts for known biases in interaction terms in PLS models
+* Continuously tested against leading PLSPM software to ensure parity of outcomes: SmartPLS (Ringle et al., 2015) and ADANCO (Henseler and Dijkstra, 2015), as well as other R packages such as semPLS (Monecke and Leisch, 2012) and matrixpls (Rönkkö, 2016)
+* _High performance, multi-core_ bootstrapping function
 
 ## Documentation
 
@@ -26,64 +37,55 @@ Demo code for use of Seminr can be found in the [seminr/demo/](https://github.co
 
 ## Installation
 
-You can install seminr from github with:
+You can install SEMinR with:
 
-```{r gh-installation, eval = FALSE}
-# install.packages("devtools")
-devtools::install_github("ISS-Analytics/seminr")
+```{r, eval=FALSE}
+install.packages("seminr")
 ```
 
 ## Usage
 
-Seminr can be used to create a plsm model, to estimate the model and to perform bootstrapping.
+Briefly, there are four steps to specifying and estimating a structural equation model using SEMinR:
 
-```{r example}
-# seminr syntax for creating measurement model
-mobi_mm <- constructs(
-  reflective("Image",        multi_items("IMAG", 1:5)),
-  reflective("Expectation",  multi_items("CUEX", 1:3)),
-  reflective("Value",        multi_items("PERV", 1:2)),
-  reflective("Satisfaction", multi_items("CUSA", 1:3))
+1 Describe measurement model for each construct and its items:
+```{r, eval=FALSE}
+# Distinguish and mix composite or reflective (common-factor) measurement models
+measurements <- constructs(
+  composite("Image",       multi_items("IMAG", 1:5), weights = mode_B),
+  composite("Expectation", multi_items("CUEX", 1:3), weights = mode_A),
+  reflective("Loyalty",    multi_items("CUSL", 1:3))
 )
+```
 
-# interaction factors must be created after the measurement model is defined
-mobi_xm <- interactions(
-  interaction_combo("Image", "Expectation"),
-  interaction_combo("Image", "Value")
+2 Specify any interactions between constructs:
+```{r, eval=FALSE}
+# Easily create orthogonalized or scaled interactions between constructs
+intxns <- interactions(
+  interaction_ortho("Image", "Expectation")
 )
+```
 
-# structural model: note that name of the interactions factor should be
-#  the names of its two main factors joined by a '.' in between.
-mobi_sm <- relationships(
-  paths(to = "Satisfaction",
-        from = c("Image", "Expectation", "Value",
-                 "Image.Expectation", "Image.Value"))
+3 Describe the structural model of causal relationships between constructs (and interactions):
+```{r, eval=FALSE}
+# Quickly create multiple paths "from" and "to" sets of constructs
+structure <- relationships(
+  paths(from = c("Image", "Expectation", "Image.Expectation"), 
+        to = "Loyalty")
 )
-
-# Load data, assemble model, and estimate using semPLS
-data("mobi", package = "semPLS")
-seminr_model <- estimate_model(data = mobi,
-                               measurement_model = mobi_mm,
-                               interactions = mobi_xm,
-                               structural_model = mobi_sm)
-
-summary(seminr_model)
-
-seminr_model_boot <- bootstrap_model(data = mobi,
-                                     measurement_model = mobi_mm,
-                                     interactions = mobi_xm,
-                                     structural_model = mobi_sm,
-                                     nboot = 200)
-
-summary(seminr_model_boot)
 ```
 
-## Testing
+4 Put the above elements together to estimate and bootstrap the model:
+```{r, eval=FALSE}
+# Dynamically compose SEM models from individual parts
+pls_model <- estimate_pls(data = mobi, measurements, intxns, structure)
+summary(pls_model)
 
-To test:
-
-```{r devtools-test, eval = FALSE}
-require(devtools)
-devtools::test()
+# Use multi-core parallel processing to speed up bootstraps
+boot_estimates <- bootstrap_model(pls_model, nboot = 1000, cores = 2)
+summary(boot_estimates)
 ```
 
+## Authors
+
+* Soumya Ray
+* Nicholas Danks

README.md

@@ -1,93 +1,86 @@
-
-<!-- README.md is generated from README.Rmd. Please edit that file -->
-SEMinR
-======
-
-The `seminr` package provides a natural syntax for researchers to describe PLS structural equation models.
-
-Documentation
--------------
-
-The vignette for Seminr can be found in the [seminr/inst/doc/](https://github.com/ISS-Analytics/seminr/blob/master/inst/doc/SEMinR.html) folder or by running the `vignette("SEMinR")` command after installation.
-
-Demo code for use of Seminr can be found in the [seminr/demo/](https://github.com/ISS-Analytics/seminr/tree/master/demo) folder or by running the `demo("seminr-contained")`, `demo("seminr-ecsi")` or `demo("seminr-interaction")` commands after installation.
-
-Installation
-------------
-
-You can install seminr from github with:
-
-``` r
-# install.packages("devtools")
-devtools::install_github("ISS-Analytics/seminr")
-```
-
-Usage
------
-
-Seminr can be used to create a plsm model, to estimate the model and to perform bootstrapping.
-
-``` r
-# seminr syntax for creating measurement model
-mobi_mm <- measure(
-  reflective("Image",        multi_items("IMAG", 1:5)),
-  reflective("Expectation",  multi_items("CUEX", 1:3)),
-  reflective("Value",        multi_items("PERV", 1:2)),
-  reflective("Satisfaction", multi_items("CUSA", 1:3))
-)
-
-# interaction factors must be created after the measurement model is defined
-mobi_xm <- interact(
-  interaction_combo("Image", "Expectation"),
-  interaction_combo("Image", "Value")
-)
-
-# structural model: note that name of the interactions factor should be
-#  the names of its two main factors joined by a '.' in between.
-mobi_sm <- structure(
-  paths(to = "Satisfaction",
-        from = c("Image", "Expectation", "Value",
-                 "Image.Expectation", "Image.Value"))
-)
-
-# Load data, assemble model, and estimate using semPLS
-data("mobi", package = "semPLS")
-seminr_model <- create_model(data = mobi,
-                             measurement_model = mobi_mm,
-                             interaction = mobi_xm,
-                             structural_model = mobi_sm)
-#> Generating the plsm model
-
-mobi_pls <- estimate_model(seminr_model, nboot = 200)
-#> Estimating model using semPLS::sempls...
-print_paths(mobi_pls)
-#>                   Satisfaction
-#> R^2                       0.60
-#> Expectation               0.47
-#> Image                     0.80
-#> Image.Expectation        -0.52
-#> Image.Value              -0.16
-#> Value                     0.43
-#>                                   Estimate Bootstrapped Estimate
-#> Expectation -> Satisfaction           0.47                  0.40
-#> Image -> Satisfaction                 0.80                  0.77
-#> Image.Expectation -> Satisfaction    -0.52                 -0.40
-#> Image.Value -> Satisfaction          -0.16                 -0.25
-#> Value -> Satisfaction                 0.43                  0.51
-#>                                   Standard Error
-#> Expectation -> Satisfaction                 0.31
-#> Image -> Satisfaction                       0.19
-#> Image.Expectation -> Satisfaction           0.51
-#> Image.Value -> Satisfaction                 0.47
-#> Value -> Satisfaction                       0.33
-```
-
-Testing
--------
-
-To test:
-
-``` r
-require(devtools)
-devtools::test()
-```
+
+<!-- README.md is generated from README.Rmd. Please edit that file -->
+SEMinR
+======
+
+SEMinR brings many advancements to creating and estimating structural equation models (SEM) using Partial Least Squares Path Modeling (PLS-PM):
+
+-   A *natural* feeling, *domain-specific* language to build and estimate structural equation models in R
+-   Uses *variance-based PLS estimation* to model both *composite* and *common-factor* constructs
+-   *High-level functions* to quickly specify interactions and complicated structural models
+
+SEMinR follows the latest best-practices in methodological literature:
+
+-   Automatically *adjusts PLS estimates to ensure consistency (PLSc)* wherever common factors are involved
+-   Ajusts for known biases in interaction terms in PLS models
+-   Continuously tested against leading PLSPM software to ensure parity of outcomes: SmartPLS (Ringle et al., 2015) and ADANCO (Henseler and Dijkstra, 2015), as well as other R packages such as semPLS (Monecke and Leisch, 2012) and matrixpls (Rönkkö, 2016)
+-   *High performance, multi-core* bootstrapping function
+
+Documentation
+-------------
+
+The vignette for Seminr can be found in the [seminr/inst/doc/](https://github.com/ISS-Analytics/seminr/blob/master/inst/doc/SEMinR.html) folder or by running the `vignette("SEMinR")` command after installation.
+
+Demo code for use of Seminr can be found in the [seminr/demo/](https://github.com/ISS-Analytics/seminr/tree/master/demo) folder or by running the `demo("seminr-contained")`, `demo("seminr-ecsi")` or `demo("seminr-interaction")` commands after installation.
+
+Installation
+------------
+
+You can install SEMinR with:
+
+``` r
+install.packages("seminr")
+```
+
+Usage
+-----
+
+Briefly, there are four steps to specifying and estimating a structural equation model using SEMinR:
+
+1 Describe measurement model for each construct and its items:
+
+``` r
+# Distinguish and mix composite or reflective (common-factor) measurement models
+measurements <- constructs(
+  composite("Image",       multi_items("IMAG", 1:5), weights = mode_B),
+  composite("Expectation", multi_items("CUEX", 1:3), weights = mode_A),
+  reflective("Loyalty",    multi_items("CUSL", 1:3))
+)
+```
+
+2 Specify any interactions between constructs:
+
+``` r
+# Easily create orthogonalized or scaled interactions between constructs
+intxns <- interactions(
+  interaction_ortho("Image", "Expectation")
+)
+```
+
+3 Describe the structural model of causal relationships between constructs (and interactions):
+
+``` r
+# Quickly create multiple paths "from" and "to" sets of constructs
+structure <- relationships(
+  paths(from = c("Image", "Expectation", "Image.Expectation"), 
+        to = "Loyalty")
+)
+```
+
+4 Put the above elements together to estimate and bootstrap the model:
+
+``` r
+# Dynamically compose SEM models from individual parts
+pls_model <- estimate_pls(data = mobi, measurements, intxns, structure)
+summary(pls_model)
+
+# Use multi-core parallel processing to speed up bootstraps
+boot_estimates <- bootstrap_model(pls_model, nboot = 1000, cores = 2)
+summary(boot_estimates)
+```
+
+Authors
+-------
+
+-   Soumya Ray
+-   Nicholas Danks

SEMinR.Rproj

@@ -19,3 +19,4 @@ BuildType: Package
 PackageUseDevtools: Yes
 PackageInstallArgs: --no-multiarch --with-keep.source
 PackageCheckArgs: --as-cran
+PackageRoxygenize: rd,collate,namespace

inst/doc/SEMinR.R

@@ -29,7 +29,7 @@ library(seminr)
 #  summary(pls_model)
 #  
 #  # Use multi-core parallel processing to speed up bootstraps
-#  boot_estimates <- bootstrap_model(pls_model, nboot = 200, cores = 4)
+#  boot_estimates <- bootstrap_model(pls_model, nboot = 2000, cores = 4)
 #  summary(boot_estimates)
 
 ## ---- eval=FALSE---------------------------------------------------------
@@ -155,7 +155,7 @@ mobi_pls <- estimate_pls(data = mobi,
 ## ------------------------------------------------------------------------
 # use 2000 bootstraps and utilize 4 parallel cores
 boot_mobi_pls <- bootstrap_model(seminr_model = mobi_pls,
-                                 nboot = 200,
+                                 nboot = 2000,
                                  cores = 2)
 
 ## ------------------------------------------------------------------------