A beginner-friendly version of the DAGitty causal-diagram editor.
▶ Try it live (free, no install): https://m-colley.github.io/dagitty/
This fork reworks the DAGitty browser GUI to make building and documenting causal diagrams (DAGs) approachable for people new to causal inference, while keeping all of the original analysis power.
What's new in the GUI
- 🎨 Modern, responsive redesign with automatic light/dark mode
- 🧭 Interactive guided tutorial for first-time users (Help → Interactive tutorial, and a one-time prompt on first visit)
- 📝 "Assumptions for your paper" — a one-click, plain-language statement of the (often untestable) assumptions your diagram requires to read a result as causal, ready to paste into a methods or limitations section
- 💬 Plain-language analysis output (e.g. “Your chosen controls block all confounding” instead of “Correctly adjusted”)
- 🩺 Diagram check — names the specific mistake and the variable involved (“M sits on a causal pathway from X to Y — controlling for it removes part of the effect you are trying to measure”): mediators, colliders, arrows drawn the wrong way round, sample selection on a common effect
- 🧰 Beginner / Advanced mode, undo / redo, empty-canvas hints, and inline help on every concept
- 💾 Never lose work — the diagram is saved in your browser and restored on the next visit;
open a saved
.dagfile from the Model menu or by dragging it onto the page - 🔗 Copy shareable link — the whole diagram travels inside the URL, so a collaborator opening it sees exactly your model. Nothing is uploaded anywhere
- 📂 Machine-readable model code plus PNG / JPEG / SVG / LaTeX / R (dagitty + ggdag) export
- 🔬 Optional Generate DAG from data — in-browser causal discovery with the PC algorithm plus
DirectLiNGAM (a port of
lingam.DirectLiNGAMv1.13.0); the suggested model is always a DAG
The diagram editing and analysis engine is the original DAGitty by Johannes Textor & Benito van der Zander (GNU GPL v2). The full upstream project description follows.
The live site is published automatically from gui/ via GitHub Pages on every push to master
(see .github/workflows/pages.yml).
This is a collection of algorithms, a GUI frontend and an R package for analyzing graphical causal models (DAGs).
The main components of the repository are:
- jslib: a JavaScript library implementing many DAG algorithms. This library underpins both the web interface and the R package, but could also be used independently, like in node.js.
- gui: HTML interface for a GUI that exposes most of the functions in the JavaScript library.
- r: R package that exposes most of the functions in the JavaScript library.
- website: The current content of dagitty.net, including a version of the GUI (which may be older than the one in gui.
- doc: LaTeX source of the dagitty PDF documentation.
Clone the repository and open the file gui/dags.html in your web browser.
Currently most functionality should work locally, but you will need an internet
connection if you want to load or save DAG models on dagitty.net.
The R package can be installed from CRAN, but this version is not updated very frequently. If you want to install the most recent version of the dagitty R package, you can:
install.packages("remotes") # unless you have it already
remotes::install_github("jtextor/dagitty/r")
If you encounter any problems installing the R package, it is probably not due to dagitty itself, but due to the package "V8" that it depends on. I may try to remove this dependency in a future version.
You can get more information on dagitty at dagitty.net and dagitty.net/learn. The R package is documented through the standard R help interface. There are also a few papers available:
-
Textor, J., van der Zander, B., Gilthorpe, M. S., Liśkiewicz, M., & Ellison, G. T. H. (2017). Robust causal inference using directed acyclic graphs: the R package ‘dagitty.’ In International Journal of Epidemiology (p. dyw341). Oxford University Press (OUP). https://doi.org/10.1093/ije/dyw341
-
Ankan, A., Wortel, I. M. N., & Textor, J. (2021). Testing Graphical Causal Models Using the R Package “dagitty.” In Current Protocols (Vol. 1, Issue 2). Wiley. https://doi.org/10.1002/cpz1.45