Retire ragas.experimental namespace (#2223)

## Summary
- Move experimental tests to appropriate locations in main test
structure
- Update build infrastructure to remove experimental references
- Remove obsolete experimental directory and scripts

Author: NirantK

.github/workflows/ci.yaml

@@ -123,10 +123,10 @@ jobs:
           # Run different test suites based on test type
           if [ "${{ matrix.test-type }}" = "full" ]; then
             # Full test suite with notebook tests
-            uv run pytest --nbmake tests/unit tests/experimental "${OPTS[@]}"
+            uv run pytest --nbmake tests/unit "${OPTS[@]}"
           else
             # Essential tests only (faster for cross-platform validation)
-            uv run pytest tests/unit tests/experimental -k "not slow" "${OPTS[@]}"
+            uv run pytest tests/unit -k "not slow" "${OPTS[@]}"
           fi
         env:
           __RAGAS_DEBUG_TRACKING: true

CLAUDE.md

@@ -88,7 +88,7 @@ make test-e2e
 
 # Direct pytest commands for more control
 uv run pytest tests/unit -k "test_name"
-uv run pytest tests/experimental -v
+uv run pytest tests/unit -v
 ```
 
 ### Documentation
@@ -161,7 +161,7 @@ The experimental features are now integrated into the main ragas package:
 To use experimental features:
 
 ```python
-from ragas.experimental import Dataset
+from ragas import Dataset
 from ragas import experiment
 from ragas.backends import get_registry
 ```

DEVELOPMENT.md

@@ -194,7 +194,7 @@ make run-ci         # Run full CI pipeline
 
 # Or use direct commands for specific tasks
 uv run pytest tests/unit          # Run core unit tests
-uv run pytest tests/experimental  # Run experimental tests
+uv run pytest tests/unit  # Run unit tests
 uv run pyright src               # Type check source code
 ```
 
@@ -278,7 +278,7 @@ make install
 uv run pytest tests/unit/test_specific.py -v
 
 # Check experimental test dependencies
-uv run pytest tests/experimental --collect-only
+uv run pytest tests/unit --collect-only
 ```
 
 #### Formatting Issues

Makefile

@@ -48,7 +48,7 @@ install: ## Install full dependencies with uv sync (backward compatible - modern
 # CODE QUALITY
 # =============================================================================
 
-.PHONY: help setup-venv install-minimal install format type check clean test test-e2e benchmarks benchmarks-docker run-ci run-ci-fast run-ci-format-check run-ci-type run-ci-tests build-docs serve-docs process-experimental-notebooks
+.PHONY: help setup-venv install-minimal install format type check clean test test-e2e benchmarks benchmarks-docker run-ci run-ci-fast run-ci-format-check run-ci-type run-ci-tests build-docs serve-docs
 format: ## Format and lint all code
 	@echo "Formatting and linting all code..."
 	@echo "(ruff format) Formatting ragas..."
@@ -84,7 +84,7 @@ benchmarks-docker: ## Run benchmarks in docker
 
 benchmarks-test: ## Run benchmarks for ragas unit tests
 	@echo "Running ragas unit tests with timing benchmarks..."
-	$(Q)uv run --active pytest --nbmake tests/unit tests/experimental --durations=0 -v $(shell if [ -n "$(k)" ]; then echo "-k $(k)"; fi)
+	$(Q)uv run --active pytest --nbmake tests/unit --durations=0 -v $(shell if [ -n "$(k)" ]; then echo "-k $(k)"; fi)
 
 # =============================================================================
 # CI/BUILD
@@ -98,7 +98,7 @@ run-ci: ## Run complete CI pipeline (mirrors GitHub CI exactly)
 	@echo "Type check..."
 	$(Q)$(MAKE) type
 	@echo "Unit tests..."
-	$(Q)__RAGAS_DEBUG_TRACKING=true RAGAS_DO_NOT_TRACK=true uv run --active pytest --nbmake tests/unit tests/experimental --dist loadfile -n auto
+	$(Q)__RAGAS_DEBUG_TRACKING=true RAGAS_DO_NOT_TRACK=true uv run --active pytest --nbmake tests/unit --dist loadfile -n auto
 	@echo "All CI checks passed!"
 
 run-ci-format-check: ## Run format check in dry-run mode (like GitHub CI)
@@ -113,23 +113,23 @@ run-ci-type: ## Run type checking (matches GitHub CI)
 
 run-ci-tests: ## Run all tests with CI options
 	@echo "Running all tests with CI options..."
-	$(Q)__RAGAS_DEBUG_TRACKING=true RAGAS_DO_NOT_TRACK=true pytest --nbmake tests/unit tests/experimental --dist loadfile -n auto
+	$(Q)__RAGAS_DEBUG_TRACKING=true RAGAS_DO_NOT_TRACK=true pytest --nbmake tests/unit --dist loadfile -n auto
 
 run-ci-fast: ## Fast CI check for quick local validation (2-3 minutes)
 	@echo "Running fast CI check for quick feedback..."
 	@echo "Format check..."
 	$(Q)uv run --active ruff format --check src tests docs --config pyproject.toml
 	$(Q)uv run --active ruff check src docs tests --config pyproject.toml
 	@echo "Core unit tests (no nbmake for speed)..."
-	$(Q)uv run --active pytest tests/unit tests/experimental --dist loadfile -n auto -x
+	$(Q)uv run --active pytest tests/unit --dist loadfile -n auto -x
 	@echo "Fast CI check completed!"
 
 clean: ## Clean all generated files
 	@echo "Cleaning all generated files..."
 	$(Q)find . -type f -name '*.py[co]' -delete -o -type d -name __pycache__ -delete
 	$(Q)rm -rf site/ docs/site/ .mypy_cache .pytest_cache .ruff_cache
-	$(Q)rm -rf dist/ build/ *.egg-info/ src/*.egg-info/ experimental/*.egg-info/
-	$(Q)rm -rf .coverage htmlcov/ .tox/ .venv/ experimental/.venv/
+	$(Q)rm -rf dist/ build/ *.egg-info/ src/*.egg-info/
+	$(Q)rm -rf .coverage htmlcov/ .tox/ .venv/
 	$(Q)find . -name '*.log' -delete
 	$(Q)find . -name '.DS_Store' -delete
 	$(Q)find . -name 'temp*' -type d -exec rm -rf {} + 2>/dev/null || true
@@ -142,11 +142,11 @@ clean: ## Clean all generated files
 
 test: ## Run all unit tests
 	@echo "Running all unit tests..."
-	$(Q)uv run --active pytest tests/unit tests/experimental $(shell if [ -n "$(k)" ]; then echo "-k $(k)"; fi)
+	$(Q)uv run --active pytest tests/unit $(shell if [ -n "$(k)" ]; then echo "-k $(k)"; fi)
 
 test-all: ## Run all unit tests (including notebooks)
 	@echo "Running all unit tests (including notebooks)..."
-	$(Q)uv run --active pytest --nbmake tests/unit tests/experimental $(shell if [ -n "$(k)" ]; then echo "-k $(k)"; fi)
+	$(Q)uv run --active pytest --nbmake tests/unit $(shell if [ -n "$(k)" ]; then echo "-k $(k)"; fi)
 
 test-e2e: ## Run all end-to-end tests
 	@echo "Running all end-to-end tests..."
@@ -156,11 +156,7 @@ test-e2e: ## Run all end-to-end tests
 # DOCUMENTATION
 # =============================================================================
 
-process-experimental-notebooks: ## Process experimental notebooks to markdown for MkDocs
-	@echo "Processing experimental notebooks..."
-	$(Q)python $(GIT_ROOT)/scripts/process_experimental_notebooks.py
-
-build-docs: process-experimental-notebooks ## Build all documentation
+build-docs: ## Build all documentation
 	@echo "Building all documentation..."
 	@echo "Converting ipynb notebooks to md files..."
 	$(Q)python $(GIT_ROOT)/docs/ipynb_to_md.py

examples/agent_evals/evals.py

@@ -1,7 +1,7 @@
 from agent import get_default_agent
 
 from ragas import experiment
-from ragas.experimental import Dataset
+from ragas import Dataset
 from ragas.metrics.numeric import numeric_metric
 from ragas.metrics.result import MetricResult
 

examples/prompt_evals/evals.py

@@ -1,7 +1,7 @@
 from prompt import run_prompt
 
 from ragas import experiment
-from ragas.experimental import Dataset
+from ragas import Dataset
 from ragas.metrics.discrete import discrete_metric
 from ragas.metrics.result import MetricResult
 

examples/rag_eval/evals.py

@@ -4,7 +4,7 @@
 from rag import default_rag_client
 
 from ragas import experiment
-from ragas.experimental import Dataset
+from ragas import Dataset
 from ragas.llms import instructor_llm_factory
 from ragas.metrics import DiscreteMetric
 

examples/workflow_eval/evals.py

@@ -4,7 +4,7 @@
 from workflow import default_workflow_client
 
 from ragas import experiment
-from ragas.experimental import Dataset
+from ragas import Dataset
 from ragas.llms import instructor_llm_factory
 from ragas.metrics import DiscreteMetric
 

pyproject.toml

@@ -12,25 +12,19 @@ dependencies = [
     "nest-asyncio",
     "appdirs",
     "diskcache>=5.6.3",
-    
-    # LangChain ecosystem
-    "langchain",
-    "langchain-core",
-    "langchain-community",
-    "langchain_openai",
-    
-    # CLI dependencies
     "typer",
     "rich",
-    
-    # LLM providers
     "openai>=1.0.0",
-    
-    # Experimental features
     "tqdm",
     "instructor",
     "gitpython",
     "pillow>=10.4.0",
+
+    # LangChain ecosystem
+    "langchain",
+    "langchain-core",
+    "langchain-community",
+    "langchain_openai",
 ]
 dynamic = ["version", "readme"]
 
@@ -117,7 +111,6 @@ preview = false
 
 [tool.pyright]
 include = ["src/ragas"]
-exclude = ["src/ragas/experimental"]
 excludeTypeshedPaths = ["@types/*"]
 pythonVersion = "3.9"
 pythonPlatform = "All"
@@ -140,10 +133,6 @@ version_file = "src/ragas/_version.py"
 addopts = "-n 0"
 asyncio_default_fixture_loop_scope = "function"
 testpaths = ["tests"]
-markers = [
-    "experimental_ci: Set of tests that will be run as part of Experimental CI",
-    "e2e: End-to-End tests for Experimental",
-]
 
 [dependency-groups]
 # Full dev dependencies with all features (used by make install)