pt.terrier.rewrite revisions - remove Axiomatic, remove terrier-prf

docs/installation.rst

@@ -91,7 +91,6 @@ These options adjust how the Terrier engine is loaded.
 
 .. autofunction:: pyterrier.terrier.set_version
 .. autofunction:: pyterrier.terrier.set_helper_version
-.. autofunction:: pyterrier.terrier.set_prf_version
 .. autofunction:: pyterrier.terrier.set_property
 .. autofunction:: pyterrier.terrier.set_properties
 .. autofunction:: pyterrier.terrier.extend_classpath

docs/pipeline_examples.md

@@ -46,7 +46,7 @@ pipe = (pt.terrier.Retriever(wikipedia_index, wmodel="BM25") >>
 
 ### RM3 Query Expansion
 
-We also provide RM3 query expansion, using an external plugin to Terrier called [terrier-prf](https://github.com/terrierteam/terrier-prf).
+We also provide RM3 query expansion.
 
 ```python
 pipe = (pt.terrier.Retriever(indexref, wmodel="BM25") >> 

docs/rewrite.rst

@@ -109,17 +109,6 @@ RM3
 References:
  - Nasreen Abdul-Jaleel, James Allan, W Bruce Croft, Fernando Diaz, Leah Larkey, Xiaoyan Li, Mark D Smucker, and Courtney Wade. UMass at TREC 2004: Novelty and HARD. In Proceedings of TREC 2004.
 
-
-AxiomaticQE
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-
-.. autoclass:: pyterrier.rewrite.AxiomaticQE
-    :members: transform 
-
-References:
- - Hui Fang, Chang Zhai.: Semantic term matching in axiomatic approaches to information retrieval. In: Proceedings of the 29th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 115–122. SIGIR 2006. ACM, New York (2006).
- - Peilin Yang and Jimmy Lin, Reproducing and Generalizing Semantic Term Matching in Axiomatic Information Retrieval. In Proceedings of ECIR 2019.
-
 Combining Query Formulations
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

pyterrier/terrier/__init__.py

@@ -1,7 +1,7 @@
 # java stuff
 from pyterrier.terrier import java
 from pyterrier.terrier._text_loader import TerrierTextLoader, terrier_text_loader
-from pyterrier.terrier.java import configure, set_version, set_helper_version, set_prf_version, extend_classpath, J, set_property, set_properties, run, version, check_version, check_helper_version
+from pyterrier.terrier.java import configure, set_version, set_helper_version, extend_classpath, J, set_property, set_properties, run, version, check_version, check_helper_version
 from pyterrier.terrier.retriever import Retriever, FeaturesRetriever, TextScorer
 from pyterrier.terrier.index_factory import IndexFactory
 from pyterrier.terrier.stemmer import TerrierStemmer
@@ -42,7 +42,7 @@ def from_dataset(*args, **kwargs):
 
 __all__ = [
     # java stuff
-    'java', 'configure', 'set_version', 'set_helper_version', 'set_prf_version', 'extend_classpath', 'J', 'version', 'check_version', 'check_helper_version',
+    'java', 'configure', 'set_version', 'set_helper_version', 'extend_classpath', 'J', 'version', 'check_version', 'check_helper_version',
 
     # retrieval
     'Retriever', 'BatchRetrieve', 'TerrierRetrieve', 'FeaturesRetriever', 'FeaturesBatchRetrieve', 'TerrierRetrieve', 'TextScorer',

pyterrier/terrier/java.py

@@ -17,7 +17,6 @@
     'helper_version': os.environ.get("TERRIER_HELPER_VERSION") or None,
     'boot_packages': [],
     'force_download': True,
-    'prf_version': '-SNAPSHOT',
 })
 
 @pt.java.before_init
@@ -30,11 +29,6 @@ def set_helper_version(version: Optional[str] = None):
     configure['helper_version'] = version
 
 
-@pt.java.before_init
-def set_prf_version(version: Optional[str] = None):
-    configure['prf_version'] = version
-
-
 class TerrierJavaInit(pt.java.JavaInitializer):
     def priority(self) -> int:
         return -10 # between pt.java.core (-100) and default (0) to load earlier than extensions
@@ -66,9 +60,6 @@ def pre_init(self, jnius_config):
         helper_jar = pt.java.mavenresolver.get_package_jar(TERRIER_PKG, 'terrier-python-helper', helper_version)
         jnius_config.add_classpath(helper_jar)
 
-        prf_jar = pt.java.mavenresolver.get_package_jar('com.github.terrierteam', 'terrier-prf', configure['prf_version'])
-        jnius_config.add_classpath(prf_jar)
-
         # This is for parallel -- it means that when re-configured in a parallel process, force_download will be False
         # and mavenresolver will use the version that was just downloaded above (not try to do it again).
         configure['force_download'] = False
@@ -140,11 +131,7 @@ def message(self):
         if "BUILD_DATE" in dir(J.Version):
             version_string += f" (build: {J.Version.BUILD_USER} {J.Version.BUILD_DATE})"
 
-        res = f"version={version_string}, helper_version={configure['helper_version']}"
-        if configure['prf_version'] is not None:
-            res += f" prf_version={configure['prf_version']}"
-
-        return res
+        return f"version={version_string}, helper_version={configure['helper_version']}"
 
     def _post_init_index(self, jnius):
         @pt.java.required
@@ -627,5 +614,4 @@ def check_helper_version(min):
     QueryResultSet = 'org.terrier.matching.QueryResultSet',
     DependenceModelPreProcess = 'org.terrier.querying.DependenceModelPreProcess',
     RM3 = 'org.terrier.querying.RM3',
-    AxiomaticQE = 'org.terrier.querying.AxiomaticQE',
 )

pyterrier/terrier/rewrite.py

@@ -390,6 +390,7 @@ def __init__(self, *args, fb_terms=10, fb_docs=3, fb_lambda=0.6, **kwargs):
             fb_docs(int): number of feedback documents to consider. Terrier's default setting is 3 feedback documents.
             fb_lambda(float): lambda in RM3, i.e. importance of relevance model viz feedback model. Defaults to 0.6.
         """
+        assert pt.terrier.check_version("5.10"), "Terrier 5.10 required"
         rm = pt.terrier.J.RM3()
         self.fb_lambda = fb_lambda
         kwargs["qeclass"] = rm
@@ -413,36 +414,6 @@ def transform(self, queries_and_docs):
         self.qe.fbDocs = self.fb_docs
         return super().transform(queries_and_docs)
 
-@pt.java.required
-class AxiomaticQE(QueryExpansion):
-    '''
-        Performs query expansion using axiomatic query expansion.
-
-        This transformer must be followed by a Terrier Retrieve() transformer.
-        The original query is saved in the `"query_0"` column, which can be restored using `pt.rewrite.reset()`.
-
-        Instance Attributes:
-         - fb_terms(int): number of feedback terms. Defaults to 10
-         - fb_docs(int): number of feedback documents. Defaults to 3
-    '''
-    def __init__(self, *args, fb_terms=10, fb_docs=3, **kwargs):
-        """
-        Args:
-            index_like: the Terrier index to use
-            fb_terms(int): number of terms to add to the query. Terrier's default setting is 10 expansion terms.
-            fb_docs(int): number of feedback documents to consider. Terrier's default setting is 3 feedback documents.
-        """
-        rm = pt.terrier.J.AxiomaticQE()
-        self.fb_terms = fb_terms
-        self.fb_docs = fb_docs
-        kwargs["qeclass"] = rm
-        super().__init__(*args, **kwargs)
-
-    def transform(self, queries_and_docs):
-        self.qe.fbTerms = self.fb_terms
-        self.qe.fbDocs = self.fb_docs
-        return super().transform(queries_and_docs)
-
 def stash_results(clear=True) -> pt.Transformer:
     """
     Stashes (saves) the current retrieved documents for each query into the column `"stashed_results_0"`.

tests/base.py

@@ -18,7 +18,6 @@ def __init__(self, *args, **kwargs):
             if terrier_helper_version is not None:
                 print("Testing with Terrier Helper version " + terrier_helper_version)
 
-            pt.terrier.set_prf_version('rm_tiebreak-SNAPSHOT')
             pt.java.set_log_level("DEBUG")
             # pt.java.add_option('-ea') can be added here to ensure that all Java assertions are met
             pt.java.init()

tests/test_rewrite.py

@@ -260,21 +260,6 @@ def test_rm3_expansion_for_query_compact_on_bm25(self):
         self.assertEqual(len(actual), 1)
         self.assertEqual(normalize_term_weights(expected), normalize_term_weights(actual.iloc[0]["query"]))
 
-    def test_axiomatic_qe_expansion_for_query_compact_on_bm25(self):
-        # just ensure that AxiomaticQE results do not change
-        expected = 'applypipeline:off compact^1.000000000'
-
-        indexref = pt.datasets.get_dataset("vaswani").get_index()
-        queriesIn = pd.DataFrame([["1", "compact"]], columns=["qid", "query"])
-
-        qe = pt.rewrite.AxiomaticQE(indexref)
-        br = pt.terrier.Retriever(indexref, wmodel='BM25')
-
-        actual = qe.transform(br.transform(queriesIn))
-
-        self.assertEqual(len(actual), 1)
-        self.assertEqual(expected, actual.iloc[0]["query"])
-
     def test_kl_qe_expansion_for_query_compact_on_bm25(self):
         # just ensure that KLQueryExpansion results do not change
         expected = 'applypipeline:off compact^1.840895333 design^0.348370740 equip^0.000000000 purpos^0.000000000 instrument^0.000000000 ferrit^0.000000000 anod^0.000000000 aircraft^0.000000000 microwav^0.000000000 sideband^0.000000000'