Improve warning message when values specified as attributes (#143)

* Improve warning message when values specified as attributes (#142)

* added test, ran black

* bugfix test

lux/core/frame.py

@@ -266,6 +266,7 @@ def set_intent(self, intent: List[Union[str, Clause]]):
         self._parse_validate_compile_intent()
 
     def _parse_validate_compile_intent(self):
+        self.maintain_metadata()
         from lux.processor.Parser import Parser
         from lux.processor.Validator import Validator
 

lux/processor/Validator.py

@@ -53,23 +53,30 @@ def validate_intent(intent: List[Clause], ldf: LuxDataFrame) -> None:
         """
 
         def validate_clause(clause):
+            warn_msg = ""
             if not (
                 (clause.attribute and clause.attribute == "?") or (clause.value and clause.value == "?")
             ):
                 if isinstance(clause.attribute, list):
                     for attr in clause.attribute:
                         if attr not in list(ldf.columns):
-                            warnings.warn(
-                                f"The input attribute '{attr}' does not exist in the DataFrame."
+                            warn_msg = (
+                                f"\n- The input attribute '{attr}' does not exist in the DataFrame."
                             )
                 else:
                     if clause.attribute != "Record":
                         # we don't value check datetime since datetime can take filter values that don't exactly match the exact TimeStamp representation
                         if clause.attribute and not is_datetime_string(clause.attribute):
                             if not clause.attribute in list(ldf.columns):
-                                warnings.warn(
-                                    f"The input attribute '{clause.attribute}' does not exist in the DataFrame."
-                                )
+                                search_val = clause.attribute
+                                match_attr = False
+                                for attr, val_list in ldf.unique_values.items():
+                                    if search_val in val_list:
+                                        match_attr = attr
+                                if match_attr:
+                                    warn_msg = f"\n- The input '{search_val}' looks like a value that belongs to the '{match_attr}' attribute. \n  Please specify the value fully, as something like {match_attr}={search_val}."
+                                else:
+                                    warn_msg = f"\n- The input attribute '{clause.attribute}' does not exist in the DataFrame. \n  Please check your input intent for typos."
                         if clause.value and clause.attribute and clause.filter_op == "=":
                             series = ldf[clause.attribute]
                             if not is_datetime_series(series):
@@ -78,15 +85,19 @@ def validate_clause(clause):
                                 else:
                                     vals = [clause.value]
                                 for val in vals:
-                                    # (not series.str.contains(val).any()):
                                     if val not in series.values:
-                                        warnings.warn(
-                                            f"The input value '{val}' does not exist for the attribute '{clause.attribute}' for the DataFrame."
-                                        )
+                                        warn_msg = f"\n- The input value '{val}' does not exist for the attribute '{clause.attribute}' for the DataFrame."
+            return warn_msg
 
+        warn_msg = ""
         for clause in intent:
             if type(clause) is list:
                 for s in clause:
-                    validate_clause(s)
+                    warn_msg += validate_clause(s)
             else:
-                validate_clause(clause)
+                warn_msg += validate_clause(clause)
+        if warn_msg != "":
+            warnings.warn(
+                "\nThe following issues are ecountered when validating the parsed intent:" + warn_msg,
+                stacklevel=2,
+            )

tests/test_error_warning.py

@@ -18,10 +18,10 @@
 from lux.vis.Vis import Vis
 
 # Test suite for checking if the expected errors and warnings are showing up correctly
-def test_context_str_error():
+def test_intent_str_error():
     df = pd.read_csv("lux/data/college.csv")
-    with pytest.raises(TypeError, match="Input intent must be a list"):
-        df.set_intent("bad string input")
+    with pytest.raises(TypeError, match="Input intent must be either a list"):
+        df.intent = "bad string input"
 
 
 def test_export_b4_widget_created():

tests/test_parser.py

@@ -14,6 +14,7 @@
 
 import pandas as pd
 import lux
+import pytest
 
 
 def test_case1():
@@ -80,9 +81,37 @@ def test_case6():
     assert ldf._intent[1].value == ["USA", "Japan", "Europe"]
 
 
-# TODO: Need to support this case
-"""
-lux.set_intent(["Horsepower","MPG","Acceleration"],"Origin")
-	lux.set_intent("Horsepower/MPG/Acceleration", "Origin")
-		--> [Clause(attr= ["Horsepower","MPG","Acceleration"], type= "attributeGroup")]
-"""
+def test_case7():
+    df = pd.read_csv("lux/data/car.csv")
+    df.intent = [["Horsepower", "MilesPerGal", "Acceleration"], "Origin"]
+    df._repr_html_()
+    assert len(df.current_vis) == 3
+
+
+def test_validator_invalid_value():
+    df = pd.read_csv("lux/data/college.csv")
+    with pytest.warns(
+        UserWarning,
+        match="The input value 'bob' does not exist for the attribute 'Region' for the DataFrame.",
+    ):
+        df.intent = ["Region=bob"]
+
+
+def test_validator_invalid_filter():
+    df = pd.read_csv("lux/data/college.csv")
+
+    with pytest.raises(KeyError, match="'New England'"):
+        with pytest.warns(
+            UserWarning,
+            match="The input 'New England' looks like a value that belongs to the 'Region' attribute.",
+        ):
+            df.intent = ["New England", "Southeast", "Far West"]
+
+
+def test_validator_invalid_attribute():
+    df = pd.read_csv("lux/data/college.csv")
+    with pytest.raises(KeyError, match="'blah'"):
+        with pytest.warns(
+            UserWarning, match="The input attribute 'blah' does not exist in the DataFrame."
+        ):
+            df.intent = ["blah"]