[GitHub] madlib pull request #240: MLP: Fix step size initialization based on learnin...
Github user asfgit closed the pull request at: https://github.com/apache/madlib/pull/240 ---
[GitHub] madlib issue #241: MiniBatch Pre-Processor: Add new module minibatch_preproc...
Github user njayaram2 commented on the issue: https://github.com/apache/madlib/pull/241 Another issue I found but forgot to mention in the review: The `__id__` column has double values instead of integers. For instance, I found values such as `0.2000` for that column in the output table. This issue also happens only when the module is used without specifying a value for the `buffer_size` param. ---
[GitHub] madlib pull request #241: MiniBatch Pre-Processor: Add new module minibatch_...
Github user njayaram2 commented on a diff in the pull request: https://github.com/apache/madlib/pull/241#discussion_r175548350 --- Diff: src/ports/postgres/modules/utilities/minibatch_preprocessing.py_in --- @@ -0,0 +1,559 @@ +# coding=utf-8 +# +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + + +""" +@file minibatch_preprocessing.py_in + +""" +from math import ceil +import plpy + +from utilities import add_postfix +from utilities import _assert +from utilities import get_seg_number +from utilities import is_platform_pg +from utilities import is_psql_numeric_type +from utilities import is_string_formatted_as_array_expression +from utilities import py_list_to_sql_string +from utilities import split_quoted_delimited_str +from utilities import _string_to_array +from utilities import validate_module_input_params +from mean_std_dev_calculator import MeanStdDevCalculator +from validate_args import get_expr_type +from validate_args import output_tbl_valid +from validate_args import _tbl_dimension_rownum + +m4_changequote(`') + +# These are readonly variables, do not modify +MINIBATCH_OUTPUT_DEPENDENT_COLNAME = "dependent_varname" +MINIBATCH_OUTPUT_INDEPENDENT_COLNAME = "independent_varname" + +class MiniBatchPreProcessor: +""" +This class is responsible for executing the main logic of mini batch +preprocessing, which packs multiple rows of selected columns from the +source table into one row based on the buffer size +""" +def __init__(self, schema_madlib, source_table, output_table, + dependent_varname, independent_varname, buffer_size, **kwargs): +self.schema_madlib = schema_madlib +self.source_table = source_table +self.output_table = output_table +self.dependent_varname = dependent_varname +self.independent_varname = independent_varname +self.buffer_size = buffer_size + +self.module_name = "minibatch_preprocessor" +self.output_standardization_table = add_postfix(self.output_table, + "_standardization") +self.output_summary_table = add_postfix(self.output_table, "_summary") +self._validate_minibatch_preprocessor_params() + +def minibatch_preprocessor(self): +# Get array expressions for both dep and indep variables from the +# MiniBatchQueryFormatter class +dependent_var_dbtype = get_expr_type(self.dependent_varname, + self.source_table) +qry_formatter = MiniBatchQueryFormatter(self.source_table) +dep_var_array_str, dep_var_classes_str = qry_formatter.\ +get_dep_var_array_and_classes(self.dependent_varname, + dependent_var_dbtype) +indep_var_array_str = qry_formatter.get_indep_var_array_str( + self.independent_varname) + +standardizer = MiniBatchStandardizer(self.schema_madlib, + self.source_table, + dep_var_array_str, + indep_var_array_str, + self.output_standardization_table) +standardize_query = standardizer.get_query_for_standardizing() + +num_rows_processed, num_missing_rows_skipped = self.\ + _get_skipped_rows_processed_count( +dep_var_array_str, +indep_var_array_str) +calculated_buffer_size = MiniBatchBufferSizeCalculator.\ + calculate_default_buffer_size( + self.buffer_size, +
[GitHub] madlib pull request #241: MiniBatch Pre-Processor: Add new module minibatch_...
Github user njayaram2 commented on a diff in the pull request: https://github.com/apache/madlib/pull/241#discussion_r175588969 --- Diff: src/ports/postgres/modules/utilities/minibatch_preprocessing.py_in --- @@ -0,0 +1,559 @@ +# coding=utf-8 +# +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + + +""" +@file minibatch_preprocessing.py_in + +""" +from math import ceil +import plpy + +from utilities import add_postfix +from utilities import _assert +from utilities import get_seg_number +from utilities import is_platform_pg +from utilities import is_psql_numeric_type +from utilities import is_string_formatted_as_array_expression +from utilities import py_list_to_sql_string +from utilities import split_quoted_delimited_str +from utilities import _string_to_array +from utilities import validate_module_input_params +from mean_std_dev_calculator import MeanStdDevCalculator +from validate_args import get_expr_type +from validate_args import output_tbl_valid +from validate_args import _tbl_dimension_rownum + +m4_changequote(`') + +# These are readonly variables, do not modify +MINIBATCH_OUTPUT_DEPENDENT_COLNAME = "dependent_varname" +MINIBATCH_OUTPUT_INDEPENDENT_COLNAME = "independent_varname" + +class MiniBatchPreProcessor: +""" +This class is responsible for executing the main logic of mini batch +preprocessing, which packs multiple rows of selected columns from the +source table into one row based on the buffer size +""" +def __init__(self, schema_madlib, source_table, output_table, + dependent_varname, independent_varname, buffer_size, **kwargs): +self.schema_madlib = schema_madlib +self.source_table = source_table +self.output_table = output_table +self.dependent_varname = dependent_varname +self.independent_varname = independent_varname +self.buffer_size = buffer_size + +self.module_name = "minibatch_preprocessor" +self.output_standardization_table = add_postfix(self.output_table, + "_standardization") +self.output_summary_table = add_postfix(self.output_table, "_summary") +self._validate_minibatch_preprocessor_params() + +def minibatch_preprocessor(self): +# Get array expressions for both dep and indep variables from the +# MiniBatchQueryFormatter class +dependent_var_dbtype = get_expr_type(self.dependent_varname, + self.source_table) +qry_formatter = MiniBatchQueryFormatter(self.source_table) +dep_var_array_str, dep_var_classes_str = qry_formatter.\ +get_dep_var_array_and_classes(self.dependent_varname, + dependent_var_dbtype) +indep_var_array_str = qry_formatter.get_indep_var_array_str( + self.independent_varname) + +standardizer = MiniBatchStandardizer(self.schema_madlib, + self.source_table, + dep_var_array_str, + indep_var_array_str, + self.output_standardization_table) +standardize_query = standardizer.get_query_for_standardizing() + +num_rows_processed, num_missing_rows_skipped = self.\ + _get_skipped_rows_processed_count( +dep_var_array_str, +indep_var_array_str) +calculated_buffer_size = MiniBatchBufferSizeCalculator.\ + calculate_default_buffer_size( + self.buffer_size, +
[GitHub] madlib pull request #241: MiniBatch Pre-Processor: Add new module minibatch_...
Github user njayaram2 commented on a diff in the pull request: https://github.com/apache/madlib/pull/241#discussion_r175593796 --- Diff: src/ports/postgres/modules/utilities/minibatch_preprocessing.py_in --- @@ -0,0 +1,559 @@ +# coding=utf-8 +# +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + + +""" +@file minibatch_preprocessing.py_in + +""" +from math import ceil +import plpy + +from utilities import add_postfix +from utilities import _assert +from utilities import get_seg_number +from utilities import is_platform_pg +from utilities import is_psql_numeric_type +from utilities import is_string_formatted_as_array_expression +from utilities import py_list_to_sql_string +from utilities import split_quoted_delimited_str +from utilities import _string_to_array +from utilities import validate_module_input_params +from mean_std_dev_calculator import MeanStdDevCalculator +from validate_args import get_expr_type +from validate_args import output_tbl_valid +from validate_args import _tbl_dimension_rownum + +m4_changequote(`') + +# These are readonly variables, do not modify +MINIBATCH_OUTPUT_DEPENDENT_COLNAME = "dependent_varname" +MINIBATCH_OUTPUT_INDEPENDENT_COLNAME = "independent_varname" + +class MiniBatchPreProcessor: +""" +This class is responsible for executing the main logic of mini batch +preprocessing, which packs multiple rows of selected columns from the +source table into one row based on the buffer size +""" +def __init__(self, schema_madlib, source_table, output_table, + dependent_varname, independent_varname, buffer_size, **kwargs): +self.schema_madlib = schema_madlib +self.source_table = source_table +self.output_table = output_table +self.dependent_varname = dependent_varname +self.independent_varname = independent_varname +self.buffer_size = buffer_size + +self.module_name = "minibatch_preprocessor" +self.output_standardization_table = add_postfix(self.output_table, + "_standardization") +self.output_summary_table = add_postfix(self.output_table, "_summary") +self._validate_minibatch_preprocessor_params() + +def minibatch_preprocessor(self): +# Get array expressions for both dep and indep variables from the +# MiniBatchQueryFormatter class +dependent_var_dbtype = get_expr_type(self.dependent_varname, + self.source_table) +qry_formatter = MiniBatchQueryFormatter(self.source_table) +dep_var_array_str, dep_var_classes_str = qry_formatter.\ +get_dep_var_array_and_classes(self.dependent_varname, + dependent_var_dbtype) +indep_var_array_str = qry_formatter.get_indep_var_array_str( + self.independent_varname) + +standardizer = MiniBatchStandardizer(self.schema_madlib, + self.source_table, + dep_var_array_str, + indep_var_array_str, + self.output_standardization_table) +standardize_query = standardizer.get_query_for_standardizing() + +num_rows_processed, num_missing_rows_skipped = self.\ + _get_skipped_rows_processed_count( +dep_var_array_str, +indep_var_array_str) +calculated_buffer_size = MiniBatchBufferSizeCalculator.\ + calculate_default_buffer_size( + self.buffer_size, +
[GitHub] madlib pull request #241: MiniBatch Pre-Processor: Add new module minibatch_...
Github user njayaram2 commented on a diff in the pull request: https://github.com/apache/madlib/pull/241#discussion_r175531202 --- Diff: src/ports/postgres/modules/utilities/minibatch_preprocessing.py_in --- @@ -0,0 +1,559 @@ +# coding=utf-8 +# +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + + +""" +@file minibatch_preprocessing.py_in + +""" +from math import ceil +import plpy + +from utilities import add_postfix +from utilities import _assert +from utilities import get_seg_number +from utilities import is_platform_pg +from utilities import is_psql_numeric_type +from utilities import is_string_formatted_as_array_expression +from utilities import py_list_to_sql_string +from utilities import split_quoted_delimited_str +from utilities import _string_to_array +from utilities import validate_module_input_params +from mean_std_dev_calculator import MeanStdDevCalculator +from validate_args import get_expr_type +from validate_args import output_tbl_valid +from validate_args import _tbl_dimension_rownum + +m4_changequote(`') + +# These are readonly variables, do not modify +MINIBATCH_OUTPUT_DEPENDENT_COLNAME = "dependent_varname" +MINIBATCH_OUTPUT_INDEPENDENT_COLNAME = "independent_varname" + +class MiniBatchPreProcessor: +""" +This class is responsible for executing the main logic of mini batch +preprocessing, which packs multiple rows of selected columns from the +source table into one row based on the buffer size +""" +def __init__(self, schema_madlib, source_table, output_table, + dependent_varname, independent_varname, buffer_size, **kwargs): +self.schema_madlib = schema_madlib +self.source_table = source_table +self.output_table = output_table +self.dependent_varname = dependent_varname +self.independent_varname = independent_varname +self.buffer_size = buffer_size + +self.module_name = "minibatch_preprocessor" +self.output_standardization_table = add_postfix(self.output_table, + "_standardization") +self.output_summary_table = add_postfix(self.output_table, "_summary") +self._validate_minibatch_preprocessor_params() + +def minibatch_preprocessor(self): +# Get array expressions for both dep and indep variables from the +# MiniBatchQueryFormatter class +dependent_var_dbtype = get_expr_type(self.dependent_varname, + self.source_table) +qry_formatter = MiniBatchQueryFormatter(self.source_table) +dep_var_array_str, dep_var_classes_str = qry_formatter.\ +get_dep_var_array_and_classes(self.dependent_varname, + dependent_var_dbtype) +indep_var_array_str = qry_formatter.get_indep_var_array_str( + self.independent_varname) + +standardizer = MiniBatchStandardizer(self.schema_madlib, + self.source_table, + dep_var_array_str, + indep_var_array_str, + self.output_standardization_table) +standardize_query = standardizer.get_query_for_standardizing() + +num_rows_processed, num_missing_rows_skipped = self.\ + _get_skipped_rows_processed_count( +dep_var_array_str, +indep_var_array_str) +calculated_buffer_size = MiniBatchBufferSizeCalculator.\ + calculate_default_buffer_size( + self.buffer_size, +
[GitHub] madlib pull request #241: MiniBatch Pre-Processor: Add new module minibatch_...
Github user njayaram2 commented on a diff in the pull request: https://github.com/apache/madlib/pull/241#discussion_r175522378 --- Diff: src/ports/postgres/modules/utilities/utilities.py_in --- @@ -794,6 +794,41 @@ def collate_plpy_result(plpy_result_rows): # -- +def validate_module_input_params(source_table, output_table, independent_varname, + dependent_varname, module_name, **kwargs): --- End diff -- How about having an optional param to deal with checking for residual output tables (summary and standardization tables). We could take a list of suffixes to check for. ---
[GitHub] madlib pull request #241: MiniBatch Pre-Processor: Add new module minibatch_...
Github user njayaram2 commented on a diff in the pull request: https://github.com/apache/madlib/pull/241#discussion_r175585050 --- Diff: src/ports/postgres/modules/utilities/minibatch_preprocessing.py_in --- @@ -0,0 +1,559 @@ +# coding=utf-8 +# +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + + +""" +@file minibatch_preprocessing.py_in + +""" +from math import ceil +import plpy + +from utilities import add_postfix +from utilities import _assert +from utilities import get_seg_number +from utilities import is_platform_pg +from utilities import is_psql_numeric_type +from utilities import is_string_formatted_as_array_expression +from utilities import py_list_to_sql_string +from utilities import split_quoted_delimited_str +from utilities import _string_to_array +from utilities import validate_module_input_params +from mean_std_dev_calculator import MeanStdDevCalculator +from validate_args import get_expr_type +from validate_args import output_tbl_valid +from validate_args import _tbl_dimension_rownum + +m4_changequote(`') + +# These are readonly variables, do not modify +MINIBATCH_OUTPUT_DEPENDENT_COLNAME = "dependent_varname" +MINIBATCH_OUTPUT_INDEPENDENT_COLNAME = "independent_varname" + +class MiniBatchPreProcessor: +""" +This class is responsible for executing the main logic of mini batch +preprocessing, which packs multiple rows of selected columns from the +source table into one row based on the buffer size +""" +def __init__(self, schema_madlib, source_table, output_table, + dependent_varname, independent_varname, buffer_size, **kwargs): +self.schema_madlib = schema_madlib +self.source_table = source_table +self.output_table = output_table +self.dependent_varname = dependent_varname +self.independent_varname = independent_varname +self.buffer_size = buffer_size + +self.module_name = "minibatch_preprocessor" +self.output_standardization_table = add_postfix(self.output_table, + "_standardization") +self.output_summary_table = add_postfix(self.output_table, "_summary") +self._validate_minibatch_preprocessor_params() + +def minibatch_preprocessor(self): +# Get array expressions for both dep and indep variables from the +# MiniBatchQueryFormatter class +dependent_var_dbtype = get_expr_type(self.dependent_varname, + self.source_table) +qry_formatter = MiniBatchQueryFormatter(self.source_table) +dep_var_array_str, dep_var_classes_str = qry_formatter.\ +get_dep_var_array_and_classes(self.dependent_varname, + dependent_var_dbtype) +indep_var_array_str = qry_formatter.get_indep_var_array_str( + self.independent_varname) + +standardizer = MiniBatchStandardizer(self.schema_madlib, + self.source_table, + dep_var_array_str, + indep_var_array_str, + self.output_standardization_table) +standardize_query = standardizer.get_query_for_standardizing() + +num_rows_processed, num_missing_rows_skipped = self.\ + _get_skipped_rows_processed_count( +dep_var_array_str, +indep_var_array_str) +calculated_buffer_size = MiniBatchBufferSizeCalculator.\ + calculate_default_buffer_size( + self.buffer_size, +
[GitHub] madlib issue #243: MLP: Add minibatch gradient descent solver
Github user asfgit commented on the issue: https://github.com/apache/madlib/pull/243 Refer to this link for build results (access rights to CI server needed): https://builds.apache.org/job/madlib-pr-build/383/ ---
[GitHub] madlib issue #245: Reduce Install Check run time
Github user asfgit commented on the issue: https://github.com/apache/madlib/pull/245 Refer to this link for build results (access rights to CI server needed): https://builds.apache.org/job/madlib-pr-build/382/ ---
[GitHub] madlib pull request #245: Reduce Install Check run time
GitHub user jingyimei opened a pull request: https://github.com/apache/madlib/pull/245 Reduce Install Check run time To reduce the total run time of install check, we looked at the top 5 modules that take longest and modified install check test cases. See each commit for details. You can merge this pull request into a Git repository by running: $ git pull https://github.com/madlib/madlib reduce_IC_run_time Alternatively you can review and apply these changes as the patch at: https://github.com/apache/madlib/pull/245.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #245 commit bfb3ef87302d80412c72f82194007f088e828974 Author: Jingyi MeiDate: 2018-03-16T00:07:44Z PCA: Refactor IC for pca_project to reduce run time To reduce the run time for pca_project, we removed calls to pca_sparse_train and pca_train. Instead, we directly insert data in the sql file. Also, we renamed some output table names for simplicity. Co-authored-by: Nikhil Kak commit a744b396cd7d21a43f7ac1bfede963a1bae318d9 Author: Jingyi Mei Date: 2018-03-16T00:45:23Z Decision Tree: Modify IC to reduce run time 1. We use a smaller array dataset as input to one of the test case, which reduced 2/3 of the decision tree IC run time. 2. Reduce n_folds from 5 to 3 in one test case Co-authored-by: Nikhil Kak commit f61bf0e88a72506e299060f3e20f062d58338ec6 Author: Jingyi Mei Date: 2018-03-16T18:41:15Z Random Forest: Clean up install check This commit reorders test cases and adds comments. Besides, it removes unnecessary casting in queries. commit 74c650ba3ef56206e86f461c00f61cb1c70fb78a Author: Jingyi Mei Date: 2018-03-16T18:44:26Z Random Foreset: Reduce install check time This commit changes num of trees from 100 to 10 in two test cases, so that the total run time of IC will be reduced. Co-authored-by: Nikhil Kak commit c5b88d55a5a9eb25c5dc9ed01f546ab31c4b8f5e Author: Jngyi Mei and Nikhil Kak Date: 2018-03-19T18:31:58Z Elastic Net: Remove cross validation test To reduce the Install Check run time for Elastic Net, we removed the test case with cross validation. Co-authored-by: Jingyi Mei Co-authored-by: Nikhil Kak ---
[GitHub] madlib issue #240: MLP: Fix step size initialization based on learning rate ...
Github user asfgit commented on the issue: https://github.com/apache/madlib/pull/240 Refer to this link for build results (access rights to CI server needed): https://builds.apache.org/job/madlib-pr-build/381/ ---