run_model: Automated Multiple Regression Modelling

In AutoModel: Automated Hierarchical Multiple Regression with Assumptions Checking

Description

Automated Multiple Regression Modelling

Usage

run_model(outcome, block1, ..., dataset, type = "gaussian",
  assumptions.check = T, outliers.check = "significant",
  transform.outcome = F)

Arguments

outcome	The dependent variable of the hierarchical model
block1	A character vector, with names of variables. The first block of independent variables.
...	A character vector, with names of variables. Subsequent blocks of independent variables.
dataset	A data frame containing variables refered to in formulas, passed to data argument of lm
type	Family argument to pass to glm. Specify "binomial" for binary logistic regression models.
assumptions.check	Boolean, if TRUE, then assumption checks are run and output is produced. If FALSE, only model summary and coefficient tables are produced.
outliers.check	Determines how many observations to display for outliers check. Default is significant observations. "All" shows all residual and Cook's D values.
transform.outcome	A boolean. If TRUE, a variable transformation of the outcome is substituted in the final model if outcome is non-normal. NOT IMPLEMENTED YET.

Details

Calls other functions to generate model objects and test them, given specified model parameters and other options. Formatted output is produced via model_output

Examples

run_model("y", c("lag.quarterly.revenue"), c("price.index", "income.level"),
dataset=freeny)

Example output

REGRESSION OUTPUT

Durbin-Watson =  2.11 p value =  0.4729 

Partial Regression plots (all relationships should be linear):

Plot of studentized residuals: uniform distibution across predicted values requiredCorrelation Matrix for model (correlation >.70 indicates severe multicollinearity)

                            y lag.quarterly.revenue price.index income.level
y                      1.0000                0.9978     -0.9895       0.9839
lag.quarterly.revenue  0.9978                1.0000     -0.9894       0.9817
price.index           -0.9895               -0.9894      1.0000      -0.9539
income.level           0.9839                0.9817     -0.9539       1.0000

Variance inflation factor (<10 desired):

lag.quarterly.revenue           price.index          income.level 
               194.85                 78.58                 45.52 

Standardized Residuals (observations > 3.00 problematic):

No significant outliers

Cook's distance (values >.2 problematic):

1963.25 
 0.8918 

Normality of standardized model residuals:  Shapiro-Wilk (p-value):  0.5586 

Model change statistics

             R    R^2 Adj R^2 SE Est. Delta R^2  F Change df1 df2 p Fch Sig
Model 1 0.9978 0.9956  0.9955  0.0212    0.9956 8360.3793   1  37     0 ***
Model 2 0.9988 0.9977  0.9975  0.0159    0.0021   15.4599   2  35     0 ***
Model 1 : y ~ lag.quarterly.revenue 
Model 2 : y ~ lag.quarterly.revenue + price.index + income.level 

Model Coefficients

   Model                  term estimate std.error statistic p.value sig
 Model 1           (Intercept)  0.04169   0.10138    0.4112  0.6833    
 Model 1 lag.quarterly.revenue  0.99827   0.01092   91.4351  0.0000 ***
 Model 2           (Intercept)  4.97077   1.24046    4.0072  0.0003 ***
 Model 2 lag.quarterly.revenue  0.37305   0.11418    3.2673  0.0024  **
 Model 2           price.index -0.81887   0.17152   -4.7742  0.0000 ***
 Model 2          income.level  0.75435   0.14454    5.2189  0.0000 ***

AutoModel documentation built on May 1, 2019, 9:14 p.m.