Updated documentation for string max_times

docs/source/automl/pipeline_search.ipynb

@@ -54,7 +54,7 @@
     {
      "data": {
       "text/plain": [
-       "<evalml.models.auto_classifier.AutoClassifier at 0x1330655d0>"
+       "<evalml.models.auto_classifier.AutoClassifier at 0x11d782750>"
       ]
      },
      "execution_count": 2,
@@ -74,7 +74,7 @@
     {
      "data": {
       "text/plain": [
-       "<evalml.models.auto_regressor.AutoRegressor at 0x133088610>"
+       "<evalml.models.auto_regressor.AutoRegressor at 0x11d792950>"
       ]
      },
      "execution_count": 3,
@@ -108,7 +108,7 @@
     {
      "data": {
       "text/plain": [
-       "<evalml.models.auto_classifier.AutoClassifier at 0x13308d610>"
+       "<evalml.models.auto_classifier.AutoClassifier at 0x11d7abf50>"
       ]
      },
      "execution_count": 4,
@@ -139,9 +139,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 5,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<evalml.models.auto_classifier.AutoClassifier at 0x11d7b3e50>"
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
    "source": [
     "from evalml.objectives import FraudCost\n",
     "\n",
@@ -171,7 +182,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -188,7 +199,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [
     {
@@ -197,7 +208,7 @@
        "[evalml.pipelines.classification.random_forest.RFClassificationPipeline]"
       ]
      },
-     "execution_count": 6,
+     "execution_count": 7,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -215,7 +226,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [
     {
@@ -226,7 +237,7 @@
        " <ModelTypes.RANDOM_FOREST: 'random_forest'>]"
       ]
      },
-     "execution_count": 7,
+     "execution_count": 8,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -237,16 +248,17 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 9,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "[<ModelTypes.RANDOM_FOREST: 'random_forest'>]"
+       "[<ModelTypes.LINEAR_MODEL: 'linear_model'>,\n",
+       " <ModelTypes.RANDOM_FOREST: 'random_forest'>]"
       ]
      },
-     "execution_count": 8,
+     "execution_count": 9,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -261,14 +273,45 @@
    "source": [
     "## Limiting Search Time\n",
     "\n",
-    "You can limit the search time by specifying a maximum number of pipelines or a maximum amount of time. EvalML won't build new pipelines after the maximum time has passed or the maximum number of pipelines have been built.\n",
+    "You can limit the search time by specifying a maximum number of pipelines and/or a maximum amount of time. EvalML won't build new pipelines after the maximum time has passed or the maximum number of pipelines have been built. If a limit is not set, then a maximum of 5 pipelines will be built. \n",
+    "\n",
+    "The maximum search time can be specified as a integer in seconds or as a string in seconds, minutes, or hours. "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<evalml.models.auto_classifier.AutoClassifier at 0x11d74c250>"
+      ]
+     },
+     "execution_count": 10,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "evalml.AutoClassifier(objective=\"f1\",\n",
+    "                      max_time=60)\n",
     "\n",
+    "evalml.AutoClassifier(objective=\"f1\",\n",
+    "                      max_time=\"1 minute\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
     "To start, EvalML samples 10 sets of hyperparameters chosen randomly for each possible pipeline. Therefore, we recommend setting `max_pipelines` at least 10 times the number of possible pipelines."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 11,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -277,16 +320,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 12,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "<evalml.models.auto_classifier.AutoClassifier at 0x1330a0250>"
+       "<evalml.models.auto_classifier.AutoClassifier at 0x11d7ce750>"
       ]
      },
-     "execution_count": 10,
+     "execution_count": 12,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -308,7 +351,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -336,7 +379,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.7.4"
+   "version": "3.7.5"
   },
   "mimetype": "text/x-python",
   "name": "python",

docs/source/changelog.rst

@@ -12,6 +12,7 @@ Changelog
     * Changes
         * Refactoring pipelines :pr:`108`
     * Documentation Changes
+        * Updated documentation to show max_time enhancements :pr:`189`
     * Testing Changes
 
 **v0.5.0 Oct. 29, 2019**

docs/source/index.ipynb

@@ -406,7 +406,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.7.4"
+   "version": "3.7.5"
   }
  },
  "nbformat": 4,

evalml/__init__.py

@@ -6,6 +6,10 @@
 # must import sklearn first
 import sklearn
 
+with warnings.catch_warnings():
+    warnings.simplefilter("ignore", DeprecationWarning)
+    import skopt
+
 import evalml.demos
 import evalml.model_types
 import evalml.objectives
@@ -20,11 +24,6 @@
 from evalml.models import AutoClassifier, AutoRegressor
 from evalml.pipelines import list_model_types, load_pipeline, save_pipeline
 
-with warnings.catch_warnings():
-    warnings.simplefilter("ignore", DeprecationWarning)
-    import skopt
-
-
 warnings.filterwarnings("ignore", category=DeprecationWarning)
 
 

evalml/models/auto_classifier.py

@@ -34,10 +34,13 @@ def __init__(self,
 
             multiclass (bool): If True, expecting multiclass data. By default: False.
 
-            max_pipelines (int): maximum number of pipelines to search
+            max_pipelines (int): Maximum number of pipelines to search. If max_pipelines and
+                max_time is not set, then max_pipelines will default to max_pipelines of 5.
 
-            max_time (int): maximum time in seconds to search for pipelines.
-                won't start new pipeline search after this duration has elapsed
+            max_time (int, str): Maximum time to search for pipelines.
+                This will not start a new pipeline search after the duration
+                has elapsed. If it is an integer, then the time will be in seconds.
+                For strings, time can be specified as seconds, minutes, or hours.
 
             model_types (list): The model types to search. By default searches over all
                 model_types. Run evalml.list_model_types("classification") to see options.

evalml/models/auto_regressor.py

@@ -29,10 +29,13 @@ def __init__(self,
         Arguments:
             objective (Object): the objective to optimize
 
-            max_pipelines (int): maximum number of pipelines to search
+            max_pipelines (int): Maximum number of pipelines to search. If max_pipelines and
+                max_time is not set, then max_pipelines will default to max_pipelines of 5.
 
-            max_time (int): maximum time in seconds to search for pipelines.
-                won't start new pipeline search after this duration has elapsed
+            max_time (int, str): Maximum time to search for pipelines.
+                This will not start a new pipeline search after the duration
+                has elapsed. If it is an integer, then the time will be in seconds.
+                For strings, time can be specified as seconds, minutes, or hours.
 
             model_types (list): The model types to search. By default searches over all
                 model_types. Run evalml.list_model_types("regression") to see options.