Introduction
A R source repository on GitHub. It provides functions for processing the outputs of Anchor to build machine learning models, and otherwise process and visualize the data.
What belongs in the repository?
Tip: It should include: R functions for processing data produced by Anchor, or data to form an input to Anchor.
Warning: It should not include: R scripts (only R functions in packages belong here!)
Packages
| Package | Description |
|---|---|
| anchor-r | General utility functions for interacting with Anchor output. |
| anchor-r-features | Processing feature-export data (CSV) from Anchor (building classifiers, plotting etc.) |
| anchor-r-experiment | Class for accessing the structured output from an Anchor experiment. |
| anchor-r-nuclei | Functions to help with nuclei segmentation pipelines. |
Code structure
- Each function is documented in its header (in the respective
R/subdirectory). - The
NAMESPACEfile specifies which functions are publicly-exposed. - R-package dependencies are listed in the
DESCRIPTIONfile and inNAMESPACE.
Installation
Warning: The packages are not on CRAN and thus cannot be imported through the usual R functions.
Before using these functions in a script elsewhere, a package must be installed locally, either:
- Use RStudio, open each project, and select the Build & Reload button in the Build panel in the top-right, or
- Use the make script (
make.bat/make.sh) in the packages’s top-level folder (after first adding your R executable to the system path).