add gridsearch tuner (#364)

* add gridsearch tuner

* add gridsearchtuner

* add gridsearchtuner

* add gridsearchtuner

* update gridsearch tuner

* update gridsearch tuner

* update gridsearch tuner

* update gridsearch tuner

* update gridsearch tuner

* update gridsearch tuner

* update gridsearch tuner

* update gridsearch and pylint

pylintrc

 # Usage:
-#       pylint --rcfile=PATH_TO_THIS_FILE PACKAGE_NAME
+#       python3 -m pylint --rcfile=PATH_TO_THIS_FILE PACKAGE_NAME
 # or
-#       pylint --rcfile=PATH_TO_THIS_FILE SOURCE_FILE.py
+#       python3 -m pylint --rcfile=PATH_TO_THIS_FILE SOURCE_FILE.py
 
 [SETTINGS]
 
 max-line-length=140
 
-max-args=5
+max-args=8
 max-locals=15
 max-statements=50
-max-attributes=7
+max-attributes=15
 
 const-naming-style=any
 
-disable=all
-
-enable=F,
-       E,
-       unreachable,
-       duplicate-key,
-       unnecessary-semicolon,
-       global-variable-not-assigned,
-       binary-op-exception,
-       bad-format-string,
-       anomalous-backslash-in-string,
-       bad-open-mode
-
-extension-pkg-whitelist=numpy
\ No newline at end of file
+disable=duplicate-code,
+        super-init-not-called
\ No newline at end of file

src/nni_manager/rest_server/restValidationSchemas.ts

@@ -61,7 +61,7 @@ export namespace ValidationSchemas {
             maxExecDuration: joi.number().min(0).required(),
             multiPhase: joi.boolean(),
             tuner: joi.object({
-                builtinTunerName: joi.string().valid('TPE', 'Random', 'Anneal', 'Evolution', 'SMAC', 'BatchTuner'),
+                builtinTunerName: joi.string().valid('TPE', 'Random', 'Anneal', 'Evolution', 'SMAC', 'BatchTuner', 'GridSearch'),
                 codeDir: joi.string(),
                 classFileName: joi.string(),
                 className: joi.string(),

src/sdk/pynni/nni/README.md

@@ -6,6 +6,7 @@ For now, NNI has supported the following tuner algorithms. Note that NNI install
  - Random Search
  - Anneal
  - Naive Evolution
+ - Grid Search
  - SMAC (to install through `nnictl`)
  - ENAS (ongoing)
  - Batch (ongoing)
@@ -46,6 +47,15 @@ Note that SMAC only supports a subset of the types in [search space spec](../../
 
 Batch allows users to simply provide several configurations (i.e., choices of hyper-parameters) for their trial code. After finishing all the configurations, the experiment is done.
 
+**Gridsearch**
+
+Gridsearch performs an exhaustive searching through a manually specified subset of the hyperparameter space defined in the searchspace file
+
+Note that the only acceptable types of search space are 'quniform', 'qloguniform' and 'choice':
+
+* Type 'choice' will select one of the options. Note that it can also be nested.
+* Type 'quniform' will receive three values [low, high, q], where [low, high] specifies a range and 'q' specifies the number of values that will be sampled evenly. It will be sampled in a way that the first sampled value is 'low', and each of the following values is (high-low)/q larger that the value in front of it.
+* Type 'qloguniform' behaves like 'quniform' except that it will first change the range to [log10(low), log10(high)] and sample and then change the sampled value back.
 
  ## 2. How to use the tuner algorithm in NNI?
 

src/sdk/pynni/nni/constants.py

@@ -25,7 +25,8 @@ ModuleName = {
     'Evolution': 'nni.evolution_tuner.evolution_tuner',
     'SMAC': 'nni.smac_tuner.smac_tuner',
     'BatchTuner': 'nni.batch_tuner.batch_tuner',
-
+    'GridSearch': 'nni.gridsearch_tuner.gridsearch_tuner',
+    
     'Medianstop': 'nni.medianstop_assessor.medianstop_assessor'
 }
 
@@ -36,6 +37,7 @@ ClassName = {
     'Evolution': 'EvolutionTuner',
     'SMAC': 'SMACTuner',
     'BatchTuner': 'BatchTuner',
+    'GridSearch': 'GridSearchTuner',
 
     'Medianstop': 'MedianstopAssessor'
 }

src/sdk/pynni/nni/gridsearch_tuner/gridsearch_tuner.py

+# Copyright (c) Microsoft Corporation
+# All rights reserved.
+#
+# MIT License
+#
+# Permission is hereby granted, free of charge,
+# to any person obtaining a copy of this software and associated
+# documentation files (the "Software"), to deal in the Software without restriction,
+# including without limitation the rights to use, copy, modify, merge, publish,
+# distribute, sublicense, and/or sell copies of the Software, and
+# to permit persons to whom the Software is furnished to do so, subject to the following conditions:
+# The above copyright notice and this permission notice shall be included
+# in all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED *AS IS*, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING
+# BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+# NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
+# DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+'''
+gridsearch_tuner.py including:
+    class GridSearchTuner
+'''
+
+import copy
+import numpy as np
+
+import nni
+from nni.tuner import Tuner
+
+TYPE = '_type'
+CHOICE = 'choice'
+VALUE = '_value'
+
+
+class GridSearchTuner(Tuner):
+    '''
+    GridSearchTuner will search all the possible configures that the user define in the searchSpace.
+    The only acceptable types of search space are 'quniform', 'qloguniform' and 'choice'
+
+    Type 'choice' will select one of the options. Note that it can also be nested.
+
+    Type 'quniform' will receive three values [low, high, q], where [low, high] specifies a range and 'q' specifies the number of values that will be sampled evenly.
+    Note that q should be at least 2.
+    It will be sampled in a way that the first sampled value is 'low', and each of the following values is (high-low)/q larger that the value in front of it.
+
+    Type 'qloguniform' behaves like 'quniform' except that it will first change the range to [log10(low), log10(high)]
+    and sample and then change the sampled value back.
+    '''
+
+    def __init__(self, optimize_mode):
+        self.count = -1
+        self.expanded_search_space = []
+
+    def json2paramater(self, ss_spec):
+        '''
+        generate all possible configs for hyperparameters from hyperparameter space.
+        ss_spec: hyperparameter space
+        '''
+        if isinstance(ss_spec, dict):
+            if '_type' in ss_spec.keys():
+                _type = ss_spec['_type']
+                _value = ss_spec['_value']
+                chosen_params = list()
+                if _type == 'choice':
+                    for value in _value:
+                        choice = self.json2paramater(value)
+                        if isinstance(choice, list):
+                            chosen_params.extend(choice)
+                        else:
+                            chosen_params.append(choice)
+                else:
+                    chosen_params = self.parse_qtype(_type, _value)
+            else:
+                chosen_params = dict()
+                for key in ss_spec.keys():
+                    chosen_params[key] = self.json2paramater(ss_spec[key])
+                return self.expand_parameters(chosen_params)
+        elif isinstance(ss_spec, list):
+            chosen_params = list()
+            for subspec in ss_spec[1:]:
+                choice = self.json2paramater(subspec)
+                if isinstance(choice, list):
+                    chosen_params.extend(choice)
+                else:
+                    chosen_params.append(choice)
+            chosen_params = list(map(lambda v: {ss_spec[0]: v}, chosen_params))
+        else:
+            chosen_params = copy.deepcopy(ss_spec)
+        return chosen_params
+
+    def _parse_quniform(self, param_value):
+        '''parse type of quniform parameter and return a list'''
+        if param_value[2] < 2:
+            raise RuntimeError("The number of values sampled (q) should be at least 2")
+        low, high, count = param_value[0], param_value[1], param_value[2]
+        interval = (high - low) / (count - 1)
+        return [float(low + interval * i) for i in range(count)]
+
+    def parse_qtype(self, param_type, param_value):
+        '''parse type of quniform or qloguniform'''
+        if param_type == 'quniform':
+            return self._parse_quniform(param_value)
+        if param_type == 'qloguniform':
+            param_value[:2] = np.log10(param_value[:2])
+            return list(np.power(10, self._parse_quniform(param_value)))
+
+        raise RuntimeError("Not supported type: %s" % param_type)
+
+    def expand_parameters(self, para):
+        '''
+        Enumerate all possible combinations of all parameters
+        para: {key1: [v11, v12, ...], key2: [v21, v22, ...], ...}
+        return: {{key1: v11, key2: v21, ...}, {key1: v11, key2: v22, ...}, ...}
+        '''
+        if len(para) == 1:
+            for key, values in para.items():
+                return list(map(lambda v: {key: v}, values))
+
+        key = list(para)[0]
+        values = para.pop(key)
+        rest_para = self.expand_parameters(para)
+        ret_para = list()
+        for val in values:
+            for config in rest_para:
+                config[key] = val
+                ret_para.append(copy.deepcopy(config))
+        return ret_para
+
+    def update_search_space(self, search_space):
+        '''
+        Check if the search space is valid and expand it: only contains 'choice' type or other types beginnning with the letter 'q'
+        '''
+        self.expanded_search_space = self.json2paramater(search_space)
+
+    def generate_parameters(self, parameter_id):
+        self.count += 1
+        if self.count > len(self.expanded_search_space)-1:
+            raise nni.NoMoreTrialError('no more parameters now.')
+        return self.expanded_search_space[self.count]
+
+    def receive_trial_result(self, parameter_id, parameters, value):
+        pass

tools/nni_cmd/config_schema.py

@@ -33,7 +33,7 @@ Optional('searchSpacePath'): os.path.exists,
 Optional('multiPhase'): bool,
 'useAnnotation': bool,
 'tuner': Or({
-    'builtinTunerName': Or('TPE', 'Random', 'Anneal', 'Evolution', 'SMAC', 'BatchTuner'),
+    'builtinTunerName': Or('TPE', 'Random', 'Anneal', 'Evolution', 'SMAC', 'BatchTuner', 'GridSearch'),
     'classArgs': {
         'optimize_mode': Or('maximize', 'minimize'),
         Optional('speed'): int