Skip to content

Home

Jump to bottom
Peter M. Dahlgren edited this page Feb 12, 2022 · 4 revisions

This wiki answers some common questions about semproducible.

How do I save my code to a file?

Use the save_code() function:

code <- semproducible(...)

save_code(code, "my_file.r")

If my_file.r already exists, the function will not continue. You need to explicitly add the overwrite=TRUE parameter to overwrite existing files:

save_code(code, "my_file.r", overwrite=TRUE)

Do I need to specify all columns?

When you want to generate code for a data frame, the default behavior is that semproducible requires you to specify which numeric columns that will be used in the covariance matrix.

Look at the iris dataset, for example:

head(iris)
  Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1          5.1         3.5          1.4         0.2  setosa
2          4.9         3.0          1.4         0.2  setosa
3          4.7         3.2          1.3         0.2  setosa
4          4.6         3.1          1.5         0.2  setosa
5          5.0         3.6          1.4         0.2  setosa
6          5.4         3.9          1.7         0.4  setosa

The column Species is a factor, which means that you will get an error since semproducible require numeric columns.

Use drop_non_numeric=TRUE to automatically drop any column that is not numeric:

semproducible(iris, drop_non_numeric=TRUE)

The following message is then displayed:

Dropped 1 non-numeric column(s): Species

How do I control the width of the code?

If you have a large data frame you want to fit on a page (e.g. journal article or appendix), you can control the physical width of the generated code. For example:

  • Set vars_per_line=2 to use two variables/values per line
  • Set digits=4 to round numbers into four decimals (the default behavior is to use as many decimals as your current R session)
semproducible(iris, drop_non_numeric=TRUE, vars_per_line=2, digits=4)

Output:

[...]

# Covariance matrix.
cov_mat <- tribble(~Sepal.Length, ~Sepal.Width, 
~Petal.Length, ~Petal.Width, 
0.6857, -0.0424, 
1.2743, 0.5163, 
-0.0424, 0.19, 
-0.3297, -0.1216, 
1.2743, -0.3297, 
3.1163, 1.2956, 
0.5163, -0.1216, 
1.2956, 0.581)

[...]

Can I use semproducible in a tidyverse pipeline?

Yes, semproducible can be used together with pipes (the %>% operator):

library(tidyverse)

iris %>%  
    select(Sepal.Length, Sepal.Width, Petal.Length) %>% 
    semproducible(formula = "Sepal.Length ~ Sepal.Width + Petal.Length") %>%
    cat()

Or directly after a lavaan object:

sem(model, data = PoliticalDemocracy) %>%
    semproducible()

How do I know that my generated code is correct?

Use the eval=TRUE argument to run and evaluate the code automatically.

code <- semproducible(iris[, 1:4], formula="Sepal.Length ~ Sepal.Width + Petal.Length", eval=TRUE)

If the code fails, you will get an error message. If the code runs successfully, semproducible will simply inform you of so and return the code.

Note: semproducible does not evaluate model fit or model convergence!

Why doesn't my generated code produce the exact same result?

The generated code will produce similar result to your original lavaan model. This is because the generated code creates a covariance matrix of your raw data, and the modeling is based on the covariance matrix.

You can control the similarity by increasing the digits argument. The higher the number, the more similar the generated result will be. However, it is likely that the value is at its maximum already.

Clone this wiki locally