Skip to content

CharlesEtienneLavoie/celflow

Repository files navigation

celflow

Lifecycle: experimental CRAN status

The goal of celflow is to provide a comprehensive set of tools and functions for optimizing the workflow of psychology research, specifically focusing on data preparation and cleaning. This package aims to streamline and simplify the process of preparing data for analysis, enabling researchers to follow good practices and adhere to established standards in the field.

By leveraging the functionality of celflow, researchers can benefit from efficient and standardized methods for data manipulation, transformation, and quality control. The package offers a collection of functions that address common challenges encountered during the data preparation phase, such as handling missing values, managing outliers, recoding variables, and conducting data checks.

With celflow, researchers can ensure the integrity and reliability of their data, reducing the potential for errors and enhancing the overall quality of their analyses. The package promotes good data management practices and facilitates the creation of reproducible research pipelines, contributing to the transparency and replicability of psychological research.

Installation

You can install the development version of celflow like so:

# If package `remotes` isn't already installed, install it with `install.packages("remotes")`
remotes::install_github("CharlesEtienneLavoie/celflow")

Example

This is a basic example which shows you how to simulate power for a Structural Equation Model (SEM) using the simulate_power function provided by celflow:

library(celflow)
library(lavaan)
#> This is lavaan 0.6-15
#> lavaan is FREE software! Please report any bugs.

# Define the lavaan model with population estimate values
modpop <- '
M ~ 0.40*X
Y ~ 0.30*M
'

# Define the lavaan model that will be tested in each simulation
mod <- '
M ~ X
Y ~ M
'
# Use the `simulate_power` function to perform power analysis
# Assume we will perform 10 simulations and generate 5 observations in each simulation
simulate_power(modpop, mod, ksim = 10, nobs = 5)
#>    lhs op rhs    est    se      z pvalue ci.lower ci.upper Parameter
#> 1    M  ~   X  0.613 0.471  1.301  0.193   -0.310    1.537     M ~ X
#> 2    Y  ~   M -0.185 0.190 -0.975  0.330   -0.557    0.187     Y ~ M
#> 3    M ~~   M  1.088 0.688  1.581  0.114   -0.261    2.437    M ~~ M
#> 4    Y ~~   Y  0.262 0.166  1.581  0.114   -0.063    0.587    Y ~~ Y
#> 5    X ~~   X  0.979 0.000     NA     NA    0.979    0.979    X ~~ X
#> 6    M  ~   X  1.050 0.265  3.964  0.000    0.531    1.569     M ~ X
#> 7    Y  ~   M  0.019 0.254  0.076  0.939   -0.479    0.518     Y ~ M
#> 8    M ~~   M  0.317 0.200  1.581  0.114   -0.076    0.709    M ~~ M
#> 9    Y ~~   Y  0.424 0.268  1.581  0.114   -0.102    0.949    Y ~~ Y
#> 10   X ~~   X  0.903 0.000     NA     NA    0.903    0.903    X ~~ X
#> 11   M  ~   X  0.377 1.049  0.360  0.719   -1.679    2.434     M ~ X
#> 12   Y  ~   M  0.371 0.350  1.062  0.288   -0.314    1.056     Y ~ M
#> 13   M ~~   M  0.931 0.589  1.581  0.114   -0.223    2.084    M ~~ M
#> 14   Y ~~   Y  0.583 0.369  1.581  0.114   -0.140    1.306    Y ~~ Y
#> 15   X ~~   X  0.169 0.000     NA     NA    0.169    0.169    X ~~ X
#> 16   M  ~   X  0.500 0.154  3.251  0.001    0.198    0.801     M ~ X
#> 17   Y  ~   M  1.887 0.134 14.087  0.000    1.625    2.150     Y ~ M
#> 18   M ~~   M  0.169 0.107  1.581  0.114   -0.040    0.378    M ~~ M
#> 19   Y ~~   Y  0.047 0.030  1.581  0.114   -0.011    0.106    Y ~~ Y
#> 20   X ~~   X  1.428 0.000     NA     NA    1.428    1.428    X ~~ X
#> 21   M  ~   X  0.103 0.255  0.405  0.686   -0.397    0.603     M ~ X
#> 22   Y  ~   M  0.099 0.347  0.285  0.776   -0.581    0.778     Y ~ M
#> 23   M ~~   M  0.385 0.244  1.581  0.114   -0.092    0.863    M ~~ M
#> 24   Y ~~   Y  0.239 0.151  1.581  0.114   -0.057    0.535    Y ~~ Y
#> 25   X ~~   X  1.184 0.000     NA     NA    1.184    1.184    X ~~ X
#> 26   M  ~   X  0.506 0.611  0.827  0.408   -0.693    1.704     M ~ X
#> 27   Y  ~   M  0.513 0.290  1.768  0.077   -0.056    1.082     Y ~ M
#> 28   M ~~   M  2.018 1.277  1.581  0.114   -0.484    4.520    M ~~ M
#> 29   Y ~~   Y  0.966 0.611  1.581  0.114   -0.232    2.164    Y ~~ Y
#> 30   X ~~   X  1.080 0.000     NA     NA    1.080    1.080    X ~~ X
#> 31   M  ~   X  0.669 0.321  2.083  0.037    0.040    1.298     M ~ X
#> 32   Y  ~   M  0.415 0.318  1.307  0.191   -0.207    1.037     Y ~ M
#> 33   M ~~   M  0.617 0.390  1.581  0.114   -0.148    1.381    M ~~ M
#> 34   Y ~~   Y  0.581 0.367  1.581  0.114   -0.139    1.301    Y ~~ Y
#> 35   X ~~   X  1.197 0.000     NA     NA    1.197    1.197    X ~~ X
#> 36   M  ~   X  0.296 0.259  1.145  0.252   -0.211    0.803     M ~ X
#> 37   Y  ~   M  1.003 0.453  2.214  0.027    0.115    1.891     Y ~ M
#> 38   M ~~   M  0.241 0.152  1.581  0.114   -0.058    0.540    M ~~ M
#> 39   Y ~~   Y  0.312 0.197  1.581  0.114   -0.075    0.699    Y ~~ Y
#> 40   X ~~   X  0.720 0.000     NA     NA    0.720    0.720    X ~~ X
#> 41   M  ~   X  0.700 0.448  1.562  0.118   -0.178    1.578     M ~ X
#> 42   Y  ~   M  0.407 0.433  0.939  0.348   -0.442    1.256     Y ~ M
#> 43   M ~~   M  0.694 0.439  1.581  0.114   -0.166    1.555    M ~~ M
#> 44   Y ~~   Y  0.970 0.613  1.581  0.114   -0.232    2.172    Y ~~ Y
#> 45   X ~~   X  0.692 0.000     NA     NA    0.692    0.692    X ~~ X
#> 46   M  ~   X -2.695 1.207 -2.234  0.025   -5.060   -0.330     M ~ X
#> 47   Y  ~   M -0.369 0.206 -1.793  0.073   -0.773    0.034     Y ~ M
#> 48   M ~~   M  0.391 0.247  1.581  0.114   -0.094    0.875    M ~~ M
#> 49   Y ~~   Y  0.165 0.105  1.581  0.114   -0.040    0.371    Y ~~ Y
#> 50   X ~~   X  0.054 0.000     NA     NA    0.054    0.054    X ~~ X

About

Helpers For Good Practices In Data Preparation

Resources

License

Unknown, MIT licenses found

Licenses found

Unknown
LICENSE
MIT
LICENSE.md

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published