Improve describe_pipeline by using pipeline's new describe method

docs/source/changelog.rst

@@ -11,6 +11,7 @@ Changelog
     * Fixes
     * Changes
         * Refactoring pipelines :pr:`108`
+        * Pulling information for describe_pipeline from pipeline's new describe method :pr:`190`
     * Documentation Changes
     * Testing Changes
 

evalml/models/auto_base.py

@@ -309,22 +309,11 @@ def describe_pipeline(self, pipeline_id, return_dict=False):
         pipeline = self.get_pipeline(pipeline_id)
         pipeline_results = self.results[pipeline_id]
 
-        self.logger.log_title("Pipeline Description")
+        pipeline.describe()
+        self.logger.log_subtitle("\nTraining")
 
-        better_string = "lower is better"
-        if pipeline.objective.greater_is_better:
-            better_string = "greater is better"
-
-        self.logger.log("Pipeline Name: %s" % pipeline.name)
-        self.logger.log("Model type: %s" % pipeline.model_type)
-        self.logger.log("Objective: %s (%s)" % (pipeline.objective.name, better_string))
-        self.logger.log("Total training time (including CV): %.1f seconds\n" % pipeline_results["training_time"])
-
-        self.logger.log_subtitle("Parameters")
-        for item in pipeline_results["parameters"].items():
-            self.logger.log("• %s: %s" % item)
-
-        self.logger.log_subtitle("\nCross Validation")
+        self.logger.log("Total training time (including CV): %.1f seconds" % pipeline_results["training_time"])
+        self.logger.log_subtitle("\nCross Validation", underline="-")
 
         if pipeline_results["high_variance_cv"]:
             self.logger.log("Warning! High variance within cross validation scores. " +

evalml/pipelines/pipeline_base.py

@@ -5,11 +5,21 @@
 
 from .components import Estimator, handle_component
 
+from evalml.model_types import ModelTypes
 from evalml.objectives import get_objective
+from evalml.problem_types import ProblemTypes
 from evalml.utils import Logger
 
 
 class PipelineBase:
+
+    model_type_dict = {ModelTypes.RANDOM_FOREST: "Random Forest",
+                       ModelTypes.XGBOOST: "XGBoost Classifier",
+                       ModelTypes.LINEAR_MODEL: "Linear Model"}
+    problem_type_dict = {ProblemTypes.BINARY: "Binary Classification",
+                         ProblemTypes.MULTICLASS: "Multiclass Classification",
+                         ProblemTypes.REGRESSION: "Regression"}
+
     def __init__(self, objective, component_list, n_jobs=-1, random_state=0):
         """Machine learning pipeline made out of transformers and a estimator.
 
@@ -76,24 +86,26 @@ def get_component(self, name):
         return next((component for component in self.component_list if component.name == name), None)
 
     def describe(self, return_dict=False):
-        """Outputs pipeline details including component parameters and cross validation information
-
-        Returns:
+        """Outputs pipeline details including component parameters
 
-            None
+        Arguments:
+            return_dict (bool): If True, return dictionary of information
+                about pipeline. Defaults to false
 
+        Returns:
+            dictionary of all component parameters if return_dict is True, else None
         """
-        title = "Pipeline: " + self.name
-        self.logger.log_title(title)
-
+        self.logger.log_title(self.name)
+        self.logger.log("Problem Types: {}".format(', '.join([self.problem_type_dict[problem_type] for problem_type in self.problem_types])))
+        self.logger.log("Model Type: {}".format(self.model_type_dict[self.model_type]))
         better_string = "lower is better"
         if self.objective.greater_is_better:
             better_string = "greater is better"
-        objective_string = "Objective: {} ({})".format(self.objective.name, better_string)
+        objective_string = "Objective to Optimize: {} ({})".format(self.objective.name, better_string)
         self.logger.log(objective_string)
 
         # Summary of steps
-        self.logger.log_subtitle("Pipeline Steps")
+        self.logger.log_subtitle("\nPipeline Steps")
         for number, component in enumerate(self.component_list, 1):
             component_string = str(number) + ". " + component.name
             self.logger.log(component_string)