Merge pull request RasaHQ#8565 from RasaHQ/merge-2.5.x-main

Merge 2.5.x into main

CHANGELOG.mdx

@@ -16,6 +16,15 @@ https://github.com/RasaHQ/rasa/tree/main/changelog/ . -->
 
 <!-- TOWNCRIER -->
 
+## [2.5.1] - 2021-04-28
+
+
+### Bugfixes
+- [#8446](https://github.com/rasahq/rasa/issues/8446): Fixed prediction for rules with multiple entities.
+- [#8545](https://github.com/rasahq/rasa/issues/8545): Mitigated Matplotlib backend issue using lazy configuration
+  and added a more explicit error message to guide users.
+
+
 ## [2.5.0] - 2021-04-12
 
 

docs/docs/setting-up-ci-cd.mdx

@@ -144,7 +144,7 @@ would be incompatible with the current production model.
 ## Example CI/CD pipelines
 
 As examples, see the CI/CD pipelines for
-[Sara](https://github.com/RasaHQ/rasa-demo/blob/master/.github/workflows/build_and_deploy.yml),
+[Sara](https://github.com/RasaHQ/rasa-demo/blob/main/.github/workflows/continuous_integration.yml),
 the Rasa assistant that you can talk to in the Rasa Docs, and
 [Carbon Bot](https://github.com/RasaHQ/carbon-bot/blob/master/.github/workflows/model_ci.yml).
 Both use [Github Actions](https://github.com/features/actions) as a CI/CD tool.

pyproject.toml

@@ -9,7 +9,7 @@ exclude = "((.eggs | .git | .pytest_cache | build | dist))"
 
 [tool.poetry]
 name = "rasa"
-version = "2.5.0"
+version = "2.5.1"
 description = "Open source machine learning framework to automate text- and voice-based conversations: NLU, dialogue management, connect to Slack, Facebook, and more - Create chatbots and voice assistants"
 authors = [ "Rasa Technologies GmbH <[email protected]>",]
 maintainers = [ "Tom Bocklisch <[email protected]>",]

rasa/core/policies/rule_policy.py

@@ -803,26 +803,33 @@ def train(
     def _does_rule_match_state(rule_state: State, conversation_state: State) -> bool:
         for state_type, rule_sub_state in rule_state.items():
             conversation_sub_state = conversation_state.get(state_type, {})
-            for key, value in rule_sub_state.items():
-                if isinstance(value, list):
-                    # json dumps and loads tuples as lists,
-                    # so we need to convert them back
-                    value = tuple(value)
+            for key, value_from_rules in rule_sub_state.items():
+                # json dumps and loads tuples as lists
+                if isinstance(value_from_rules, list):
+                    # sort values before comparing
+                    # `sorted` returns a list
+                    value_from_rules = sorted(value_from_rules)
+
+                value_from_conversation = conversation_sub_state.get(key)
+                if isinstance(value_from_conversation, tuple):
+                    # sort values before comparing
+                    # `sorted` returns a list
+                    value_from_conversation = sorted(value_from_conversation)
 
                 if (
                     # value should be set, therefore
                     # check whether it is the same as in the state
-                    value
-                    and value != SHOULD_NOT_BE_SET
-                    and conversation_sub_state.get(key) != value
+                    value_from_rules
+                    and value_from_rules != SHOULD_NOT_BE_SET
+                    and value_from_conversation != value_from_rules
                 ) or (
                     # value shouldn't be set, therefore
                     # it should be None or non existent in the state
-                    value == SHOULD_NOT_BE_SET
-                    and conversation_sub_state.get(key)
+                    value_from_rules == SHOULD_NOT_BE_SET
+                    and value_from_conversation
                     # during training `SHOULD_NOT_BE_SET` is provided. Hence, we also
                     # have to check for the value of the slot state
-                    and conversation_sub_state.get(key) != SHOULD_NOT_BE_SET
+                    and value_from_conversation != SHOULD_NOT_BE_SET
                 ):
                     return False
 

rasa/utils/plotting.py

@@ -1,9 +1,10 @@
 import logging
 import itertools
 import os
+from functools import wraps
 
 import numpy as np
-from typing import List, Text, Optional, Union, Any
+from typing import Any, Callable, List, Optional, Text, TypeVar, Union
 import matplotlib
 from matplotlib.ticker import FormatStrFormatter
 
@@ -14,10 +15,20 @@
 
 
 def _fix_matplotlib_backend() -> None:
-    """Tries to fix a broken matplotlib backend..."""
+    """Tries to fix a broken matplotlib backend."""
+    try:
+        backend = matplotlib.get_backend()
+    except Exception:  # skipcq:PYL-W0703
+        logger.error(
+            "Cannot retrieve Matplotlib backend, likely due to a compatibility "
+            "issue with system dependencies. Please refer to the documentation: "
+            "https://matplotlib.org/stable/tutorials/introductory/usage.html#backends"
+        )
+        raise
+
     # At first, matplotlib will be initialized with default OS-specific
     # available backend
-    if matplotlib.get_backend() == "TkAgg":
+    if backend == "TkAgg":
         try:
             # on OSX sometimes the tkinter package is broken and can't be imported.
             # we'll try to import it and if it fails we will use a different backend
@@ -27,7 +38,7 @@ def _fix_matplotlib_backend() -> None:
             matplotlib.use("agg")
 
     # if no backend is set by default, we'll try to set it up manually
-    elif matplotlib.get_backend() is None:  # pragma: no cover
+    elif backend is None:  # pragma: no cover
         try:
             # If the `tkinter` package is available, we can use the `TkAgg` backend
             import tkinter  # noqa: 401
@@ -39,10 +50,27 @@ def _fix_matplotlib_backend() -> None:
             matplotlib.use("agg")
 
 
-# we call the fix as soon as this package gets imported
-_fix_matplotlib_backend()
+ReturnType = TypeVar("ReturnType")
+FuncType = Callable[..., ReturnType]
+_MATPLOTLIB_BACKEND_FIXED = False
+
+
+def _needs_matplotlib_backend(func: FuncType) -> FuncType:
+    """Decorator to fix matplotlib backend before calling a function."""
+
+    @wraps(func)
+    def inner(*args: Any, **kwargs: Any) -> ReturnType:
+        """Replacement function that fixes matplotlib backend."""
+        global _MATPLOTLIB_BACKEND_FIXED
+        if not _MATPLOTLIB_BACKEND_FIXED:
+            _fix_matplotlib_backend()
+            _MATPLOTLIB_BACKEND_FIXED = True
+        return func(*args, **kwargs)
+
+    return inner
 
 
+@_needs_matplotlib_backend
 def plot_confusion_matrix(
     confusion_matrix: np.ndarray,
     classes: Union[np.ndarray, List[Text]],
@@ -117,6 +145,7 @@ def plot_confusion_matrix(
         fig.savefig(output_file, bbox_inches="tight")
 
 
+@_needs_matplotlib_backend
 def plot_histogram(
     hist_data: List[List[float]], title: Text, output_file: Optional[Text] = None
 ) -> None:
@@ -217,6 +246,7 @@ def plot_histogram(
         fig.savefig(output_file, bbox_inches="tight")
 
 
+@_needs_matplotlib_backend
 def plot_curve(
     output_directory: Text,
     number_of_examples: List[int],

rasa/version.py

@@ -1,3 +1,3 @@
 # this file will automatically be changed,
 # do not add anything but the version number here!
-__version__ = "2.5.0"
+__version__ = "2.5.1"

tests/core/policies/test_rule_policy.py

@@ -27,7 +27,7 @@
     RULE_ONLY_SLOTS,
     RULE_ONLY_LOOPS,
 )
-from rasa.shared.nlu.constants import TEXT, INTENT, ACTION_NAME
+from rasa.shared.nlu.constants import TEXT, INTENT, ACTION_NAME, ENTITY_ATTRIBUTE_TYPE
 from rasa.shared.core.domain import Domain
 from rasa.shared.core.events import (
     ActionExecuted,
@@ -2101,9 +2101,7 @@ def test_hide_rule_turn():
         ],
     )
     policy = RulePolicy()
-    policy.train(
-        [GREET_RULE, chitchat_story], domain, RegexInterpreter(),
-    )
+    policy.train([GREET_RULE, chitchat_story], domain, RegexInterpreter())
 
     conversation_events = [
         ActionExecuted(ACTION_LISTEN_NAME),
@@ -2664,9 +2662,7 @@ def test_remove_action_listen_prediction_if_contradicts_with_story():
         ],
     )
     policy = RulePolicy()
-    policy.train(
-        [rule, story], domain, RegexInterpreter(),
-    )
+    policy.train([rule, story], domain, RegexInterpreter())
     prediction_source = [{PREVIOUS_ACTION: {ACTION_NAME: utter_1}}]
     key = policy._create_feature_key(prediction_source)
     assert key not in policy.lookup[RULES]
@@ -2718,9 +2714,7 @@ def test_keep_action_listen_prediction_after_predictable_action():
     # prediction of action_listen should only be removed if it occurs after the first
     # action (unpredictable)
     with pytest.raises(InvalidRule):
-        policy.train(
-            [rule, story], domain, RegexInterpreter(),
-        )
+        policy.train([rule, story], domain, RegexInterpreter())
 
 
 def test_keep_action_listen_prediction_if_last_prediction():
@@ -2762,9 +2756,7 @@ def test_keep_action_listen_prediction_if_last_prediction():
     policy = RulePolicy()
     # prediction of action_listen should only be removed if it's not the last prediction
     with pytest.raises(InvalidRule):
-        policy.train(
-            [rule, story], domain, RegexInterpreter(),
-        )
+        policy.train([rule, story], domain, RegexInterpreter())
 
 
 def test_keep_action_listen_prediction_if_contradicts_with_rule():
@@ -2807,9 +2799,7 @@ def test_keep_action_listen_prediction_if_contradicts_with_rule():
     )
     policy = RulePolicy()
     with pytest.raises(InvalidRule):
-        policy.train(
-            [rule, other_rule], domain, RegexInterpreter(),
-        )
+        policy.train([rule, other_rule], domain, RegexInterpreter())
 
 
 def test_raise_contradiction_if_rule_contradicts_with_story():
@@ -2850,6 +2840,66 @@ def test_raise_contradiction_if_rule_contradicts_with_story():
     )
     policy = RulePolicy()
     with pytest.raises(InvalidRule):
-        policy.train(
-            [rule, story], domain, RegexInterpreter(),
-        )
+        policy.train([rule, story], domain, RegexInterpreter())
+
+
+def test_rule_with_multiple_entities():
+    intent_1 = "intent_1"
+    entity_1 = "entity_1"
+    entity_2 = "entity_2"
+    utter_1 = "utter_1"
+    domain = Domain.from_yaml(
+        f"""
+        version: "2.0"
+        intents:
+        - {intent_1}
+        entities:
+        - {entity_1}
+        - {entity_2}
+        actions:
+        - {utter_1}
+        """
+    )
+
+    rule = TrackerWithCachedStates.from_events(
+        "rule without action_listen",
+        domain=domain,
+        slots=domain.slots,
+        evts=[
+            ActionExecuted(RULE_SNIPPET_ACTION_NAME),
+            ActionExecuted(ACTION_LISTEN_NAME),
+            UserUttered(
+                intent={"name": intent_1},
+                entities=[
+                    {ENTITY_ATTRIBUTE_TYPE: entity_1},
+                    {ENTITY_ATTRIBUTE_TYPE: entity_2},
+                ],
+            ),
+            ActionExecuted(utter_1),
+            ActionExecuted(ACTION_LISTEN_NAME),
+        ],
+        is_rule_tracker=True,
+    )
+    policy = RulePolicy()
+    policy.train([rule], domain, RegexInterpreter())
+
+    # the order of entities in the entities list doesn't matter for prediction
+    conversation_events = [
+        ActionExecuted(ACTION_LISTEN_NAME),
+        UserUttered(
+            "haha",
+            intent={"name": intent_1},
+            entities=[
+                {ENTITY_ATTRIBUTE_TYPE: entity_2},
+                {ENTITY_ATTRIBUTE_TYPE: entity_1},
+            ],
+        ),
+    ]
+    prediction = policy.predict_action_probabilities(
+        DialogueStateTracker.from_events(
+            "casd", evts=conversation_events, slots=domain.slots
+        ),
+        domain,
+        RegexInterpreter(),
+    )
+    assert_predicted_action(prediction, domain, utter_1)