17  Storing Outputs and Metadata Across Sessions

ggsem enables users to store outputs and metadata in two different ways.

1. Save Output Tables as CSVs

You can save output tables from the app in csv formats. These can be re-uploaded in the app directly, and they can also be converted to plot outputs using the function csv_to_ggplot(). However, it does not save the history of the entire session from the app.

Example of this method is in Chapter 3 in section 2, as pasted here:

library(ggsem)

points <- read.csv('batch/points_batch.csv') 
lines <- read.csv('batch/lines_batch.csv') 
annotations <- read.csv('batch/annotations_batch.csv') 

ggsem_data <- list(points, lines, annotations) # Put them in a list, any order is fine

plot3 <- csv_to_ggplot(ggsem_data)

save_figure('plot3.png', plot3)

2. Save Metadata to Capture Entire Workflow History

Additionally, you can save the entire metadata from one session of the app by clicking Capture Complete Workflow, and then Download Complete Workflow. This metadata can be converted into a ggplot object using metadata_to_ggplot().

library(tidyverse)
library(ggsem)

metadata <- readRDS('ggsem_metadata.rds') # load metadata from the app
plot1 <- metadata_to_ggplot(metadata) # convert RDS into a list of data frames (output tables)

save_figure('plot1.png', plot1)

Figure 1. Save and load metadata (RDS) file

If you want to manually edit data frames containing information about graphics, you can use ggsem_silent() instead, which returns a list of output tables. ggsem_silent() re-runs the whole session’s metadata and reproduces the same diagram outputs. The output of ggsem_silent() can be converted into ggplot object using csv_to_ggplot().

library(ggsem)

metadata <- readRDS('ggsem_metadata.rds') # load metadata from the app
ggsem_data <- ggsem_silent(metadata) # edit ggsem_data 
plot1 <- csv_to_ggplot(ggsem_data) # convert RDS into a list of data frames (output tables)

save_figure('plot1.png', plot1)

Save Current Work

Capture Complete Workflow:

  • Records all aspects of your current session including:

    • All group definitions and configurations

    • Statistical models and parameter estimates

    • Imported data files and datasets

    • Visual elements, layouts, and aesthetic settings

    • Current diagram state and positioning

Download Complete Workflow:

  • Exports workflow as a portable .rds file

  • Enables figure reproduction without relaunching the Shiny app

  • Preserves all analytical and visual components for future use

  • Supports collaborative sharing and publication reproducibility

Continue Previous Work

Upload Workflow to Continue:

  • Load previously saved .rds workflow files

  • Restore complete analysis environment exactly as saved

  • Resume work from any previous session state

  • Maintain all model specifications, data, and visual customizations

  • Workflow can be replayed using Redo and Undo buttons

Metadata Status

Workflow Summary:

  • Displays current session metadata including:

    • Number of groups for SEM and networks

    • Session click counts and other information