From e506957c8971ecdc92ab83edf0916584c58ac074 Mon Sep 17 00:00:00 2001 From: Kevin Maik Jablonka Date: Mon, 12 Aug 2024 15:25:44 -0700 Subject: [PATCH] update docs on identifier wrapping --- docs/api/sampler.md | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/docs/api/sampler.md b/docs/api/sampler.md index e289d50a6..372a98782 100644 --- a/docs/api/sampler.md +++ b/docs/api/sampler.md @@ -119,7 +119,12 @@ print(mc_result) ## Notes -- The `TemplateSampler` class now supports wrapping of identifiers with tags when the `wrap_identifiers` option is enabled in the configuration. +- The `TemplateSampler` class supports various types of templates, including those with multiple choice questions. +- Class balancing can be useful for creating balanced datasets for machine learning tasks. +- The sampler can handle both categorical and continuous data types, with proper formatting for continuous values. +- Custom sampling functions can be provided for more control over how values are selected from multiple options. + +- The `TemplateSampler` class supports wrapping of identifiers with tags when the `wrap_identifiers` option is enabled in the configuration. - Wrapped identifiers use the format `[BEGIN_IDENTIFIER_TYPE]value[END_IDENTIFIER_TYPE]`. - Identifier types are based on the `IdentifierEnum` class, which includes common chemical identifiers like SMILES, InChI, and others.