FreesearchR/vignettes/FreesearchR.Rmd

---
title: "FreesearchR"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{FreesearchR}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE,eval = FALSE)
library(FreesearchR)
```

# Getting started with ***FreesearchR***

Below is a simple walk-trough and basic descriptions on the different features of the ***FreesearchR*** app.

## Launching

The easiest way to get started is to launch [the onlie version of the app (click this link)](https://app.freesearchr.org/). Please be aware not to upload sensitive data in this version as data security can not be guaranteed in this online environment. The app can easily be run from *R* on your own computer by running the code below ([read more on running locally here](https://agdamsbo.github.io/FreesearchR/#run-locally-on-your-own-machine)):

```{r}
require("pak")
pak::pak("agdamsbo/FreesearchR")
library(FreesearchR)
FreesearchR::launch_FreesearchR()
```

As a small note, a standalone Windows app version is on the drawing board as well, but no time frame is currently available.

## Get started

Once in the app, get started by loading your data. You have three options available for importing data: file upload, REDCap server export and local or sample data.

After choosing a data source nad importing data, you can preview the basic data structure and missing observations, set a threshold to filter data by completeness and further manually specify variables to include for analyses.

### File upload

Several data file formats are supported for easy import (csv, txt, xls(x), ods, rds, dta). If importing workbooks (xls(x) or ods), you are prompted to specify sheet(s) to import. If choosing multiple sheets, these are automatically merged by common variable(s), so please make sure that key/ID variables are correctly named identically.

### REDCap server export

Export data directly from a REDCap server. You need to first generate an API-token ([see these instruction](https://confluence.research.cchmc.org/pages/viewpage.action?pageId=50987698)) in REDCap. Make sure you have the necessary rights to do so.

Please don't store the API-key on your device unless encrypted or in a keyring, as this may compromise data safety. Log in to your REDCap server and retrieve the token when needed.

Type the correct web address of your REDCap server.

The module will validate the information and you can click "Connect".

This will unfold options to preview your data dictionary (the main database metadata), choose fields/variables to download as well as filtering options.

### Local or sample data

When opening the online hosted app, you can load some sample data to try out the app. When running the app locally from *R* on your own computer, you will find all data frames loaded in your environment here. This extends the possible uses of this app to allow for quick and easy data insights and code generation.

## Prepare

This is the panel to prepare data for evaluation and analyses and get a good overview of your data, check data is classed and formatted correctly, perform simple modifications and filter data. 

### Summary

Here, the data variables can be inspected with a simple visualisation and a few key measures. Also, data filtering is available at two levels: 

   - Data type filtering allows to filter by variable [data type](https://agdamsbo.github.io/FreesearchR/articles/data-types.html)
   
   - Observations level filtering allow to filter data by variable

### Modify

Re-class, rename, and relabel variables. Subset data, create new variables and reorder factor levels. Also, compare the modified dataset to the original and restore the original data.

## Evaluate

This panel allows for basic data evaluation.

### Characteristics

Create a classical baseline characteristics table with optional data stratification and comparisons.

### Correlation matrix

Visualise variable correlations and get suggestions to exclude highly correlated variables.

## Visuals

There are a number of plotting options to visualise different aspects of the data.

Below are the available plot types listed.

```{r echo = FALSE, eval = TRUE}
c("continuous", "dichotomous", "categorical") |> 
  lapply(\(.x){
    dplyr::bind_cols(
      dplyr::tibble("Data type"=.x),
      supported_plots() |> 
      lapply(\(.y){
        if (.x %in% .y$primary.type){
          .y[c("descr","note")]|> dplyr::bind_cols()
        }
})|> 
  dplyr::bind_rows() |> 
  setNames(c("Plot type","Description")))
  }) |> 
  dplyr::bind_rows() |> 
  # toastui::datagrid(filters=TRUE,theme="striped") |> 
  knitr::kable()
```

Export the plots directly from the sidebar with easily adjusted plot dimensions for your next publication.

Also copy the code to generate the plot in your own R-environment and fine tune all the small details.

## Regression

This section is only intended for very simple explorative analyses and as a proof-of-concept for now. If you are doing complex regression analyses you should probably just write the code yourself.

Below are the available regression types listed.

```{r echo = FALSE, eval = TRUE}
c("continuous", "dichotomous", "categorical") |> 
  lapply(\(.x){
    dplyr::bind_cols(
      dplyr::tibble("Data type"=.x),
      supported_functions()|> 
      lapply(\(.y){
        if (.x %in% .y$out.type){
          .y[c("descr","fun","design")]|> dplyr::bind_cols()
        }
})|> 
  dplyr::bind_rows() |> 
  setNames(c("Regression model","Function","Study design")))
  }) |> 
  dplyr::bind_rows() |> 
  # toastui::datagrid(filters=TRUE,theme="striped") |> 
  knitr::kable()
```

### Table

Generate simple regression models and get the results in a nice table. This will also be included in the exported report.

This will generate a combined table with both univariate regression model results for each included variable and a multivariate model with all variables included for explorative analyses.

### Plots

Plot the coefficients from the regression models in a forest plot. Choose which model(s) to include.

### Model checks

Check model assumptions visually. Supported checks can be chosen.

## Download

### Report

Download a nice report with baseline characteristics and regression model results. Choose between MS Word or LibreOffice format.

### Data

Export the modified dataset in different formats.

### Code

See all the code snippets from the different steps in your data evaluation.