ImPlot is an immediate mode, GPU accelerated plotting library for [Dear ImGui](https://github.com/ocornut/imgui). It aims to provide a first-class API that ImGui fans will love. ImPlot is well suited for visualizing program data in real-time or creating interactive plots, and requires minimal code to integrate. Just like ImGui, it does not burden the end user with GUI state management, avoids STL containers and C++ headers, and has no external dependencies except for ImGui itself.
The API is used just like any other ImGui `BeginX`/`EndX` pair. First, start a new plot with `ImPlot::BeginPlot()`. Next, plot as many items as you want with the provided `PlotX` functions (e.g. `PlotLine()`, `PlotBars()`, `PlotScatter()`, etc). Finally, wrap things up with a call to `ImPlot::EndPlot()`. That's it!
A comprehensive example of ImPlot's features can be found in `implot_demo.h`. Add this file to your sources and call `ImPlot::ShowDemoWindow()` somewhere in your update loop. You are encouraged to use this file as a reference when needing to implement various plot types. The demo is always updated to show new plot types and features as they are added, so check back with each release!
An online version of the demo is hosted [here](https://traineq.org/implot_demo/src/implot_demo.html). You can view the plots and the source code that generated them. Note that this demo may not always be up to date and is not as performant as a desktop implementation, but it should give you a general taste of what's possible with ImPlot. Special thanks to [pthom](https://github.com/pthom) for creating and hosting this!
More sophisticated demos requiring lengthier code and/or third-party libraries can be found in a separate repository: [implot_demos](https://github.com/epezent/implot_demos). Here, you will find advanced signal processing and ImPlot usage in action. Please read the `Contributing` section of that repository if you have an idea for a new demo!
0) Set up an [ImGui](https://github.com/ocornut/imgui) environment if you don't already have one.
1) Add `implot.h`, `implot_internal.h`, `implot.cpp`, `implot_items.cpp` and optionally `implot_demo.cpp` to your sources. Alternatively, you can get ImPlot using [vcpkg](https://github.com/microsoft/vcpkg/tree/master/ports/implot).
If you want to test ImPlot quickly, consider trying [mahi-gui](https://github.com/mahilab/mahi-gui), which bundles ImGui, ImPlot, and several other packages for you.
Dear ImGui uses **16-bit indexing by default**, so high-density ImPlot widgets like `ImPlot::PlotHeatmap()` may produce too many vertices into `ImDrawList`, which causes an assertion failure and will result in data truncation and/or visual glitches. Therefore, it is **HIGHLY** recommended that you EITHER:
- **Option 1:** Enable 32-bit indices by uncommenting `#define ImDrawIdx unsigned int` in your ImGui [`imconfig.h`](https://github.com/ocornut/imgui/blob/master/imconfig.h#L89) file.
- **Option 2:** Handle the `ImGuiBackendFlags_RendererHasVtxOffset` flag in your renderer if you must use 16-bit indices. Many of the default ImGui rendering backends already support `ImGuiBackendFlags_RendererHasVtxOffset`. Refer to [this issue](https://github.com/ocornut/imgui/issues/2591) for more information.
A: ImGui is an incredibly powerful tool for rapid prototyping and development, but provides only limited mechanisms for data visualization. Two dimensional plots are ubiquitous and useful to almost any application. Being able to visualize your data in real-time will give you insight and better understanding of your application.
A: If you're looking to generate publication quality plots and/or export plots to a file, ImPlot is NOT the library for you! ImPlot is geared toward plotting application data at realtime speeds with high levels of interactivity. ImPlot does its best to create pretty plots (indeed, there are quite a few styling options available), but it will always favor function over form.
A: The API is thoroughly commented in `implot.h`, and the demo in `implot_demo.cpp` should be more than enough to get you started. Also take a look at the [implot_demos](https://github.com/epezent/implot_demos) repository.
A: Yes, within reason. You can plot tens to hundreds of thousands of points without issue, but don't expect millions to be a buttery smooth experience. That said, you can always downsample extremely large datasets by telling ImPlot to stride your data at larger intervals if needed. Also try the experimental `backends` branch which aims to provide GPU acceleration support.
`float`, `double`, `int8`, `uint8`, `int16`, `uint16`, `int32`, `uint32`, `int64`, `uint64`. Arrays of custom structs or classes (e.g. `Vector2f` or similar) are easily passed to ImPlot functions using the built in striding features (see `implot.h` for documentation), and many plotters provide a "getter" overload which accepts data generating callbacks.
A: Yes. Data colormaps and various styling colors and variables can be pushed/popped or modified permanently on startup. Three default styles are available, as well as an automatic style that attempts to match you ImGui style.
A: Maybe. Check the demo, gallery, or Announcements ([2020](https://github.com/epezent/implot/issues/48)/[2021](https://github.com/epezent/implot/issues/168))to see if your desired plot type is shown. If not, consider submitting an issue or better yet, a PR!
A: By default, no anti-aliasing is done on line plots for performance gains. If you use at least 4x MSAA, then you likely won't even notice. However, you can enable software AA per-plot with the `ImPlotFlags_AntiAliased` flag, or globally with `ImPlot::GetStyle().AntiAliasedLines = true;`.
A: Not currently. Use your OS's screen capturing mechanisms if you need to capture a plot. ImPlot is not suitable for rendering publication quality plots; it is only intended to be used as a visualization tool. Post-process your data with MATLAB or matplotlib for these purposes.
A: Like ImGui, it is recommended that you compile and link ImPlot as a *static* library or directly as a part of your sources. However, if you must and are compiling ImPlot and ImGui as separate DLLs, make sure you set the current *ImGui* context with `ImPlot::SetImGuiContext(ImGuiContext* ctx)`. This ensures that global ImGui variables are correctly shared across the DLL boundary.
A: Yes, you can use the generated C binding, [cimplot](https://github.com/cimgui/cimplot) with most high level languages. [DearPyGui](https://github.com/hoffstadt/DearPyGui) provides a Python wrapper, among other things. [imgui-java](https://github.com/SpaiR/imgui-java) provides bindings for Java. A Rust binding, [implot-rs](https://github.com/4bb4/implot-rs), is currently in the works. An example using Emscripten can be found [here](https://github.com/pthom/implot_demo).
