Add erfinv operator for calculating inverse error function #13811

rondogency · 2019-01-09T05:22:22Z

Description

Add erfinv math operator. This is an extension of the previous error function implementation #13229.

Checklist

Essentials

Please feel free to remove inapplicable items for your PR.

The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant JIRA issue created (except PRs with tiny changes)
Changes are complete (i.e. I finished coding on this PR)
All changes have test coverage:
Unit tests are added for small changes to verify correctness (e.g. adding a new operator)
Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore)
Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL)
Code is well-documented:
For user-facing API changes, API doc string has been updated.
For new C++ functions in header files, their functionalities and arguments are documented.
For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable
Check the API doc at http:https://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html
To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change

Changes

None

Comments

None

This reverts commit 64e9f1a.

src/operator/mshadow_op.h

anirudhacharya · 2019-01-09T18:23:38Z

@mxnet-label-bot add [pr-awaiting-review]

src/operator/mshadow_op.h

src/operator/contrib/erfinv-inl.h

eric-haibin-lin

Coud you add this file to https://github.com/apache/incubator-mxnet/blob/master/tests/nightly/apache_rat_license_check/rat-excludes to pass license check?

rondogency · 2019-01-19T04:36:00Z

@eric-haibin-lin the unix gpu failed on a dropout_perf test. Have you even seen it before? I am a bit struggling to find the root cause of it.

eric-haibin-lin · 2019-01-19T05:16:55Z

The failure does not seem related. I've triggered the test again to see if it passes this time

src/operator/contrib/erfinv-inl.h

eric-haibin-lin

Looks good pending one minor comment

tools/license_header.py

) * add default behaviour for argmax * prototype of erfvin * add test * gpu support * Revert "add default behaviour for argmax" This reverts commit 64e9f1a. * move erfinv to contrib * edit copyright * remove atof * use std and update license * add license exclude file * fix per eric's comments * change license header

TaoLv · 2019-02-21T14:36:10Z

@rondogency Could you take a look at those warnings introduced by this PR? If possible we need get rid of them.

src/operator/tensor/./.././contrib/erfinv-inl.h(79): warning: floating-point value does not fit in required integral type
          detected during:
            instantiation of "DType mxnet::op::mshadow_op::erfinv::Map(DType) [with DType=uint8_t]"
src/operator/tensor/./../mxnet_op.h(479): here
            instantiation of "void mxnet::op::mxnet_op::op_with_req<OP, req>::Map(mxnet::index_t, DType *, const DType *) [with OP=mxnet::op::mshadow_op::erfinv, req=1, DType=uint8_t]"
src/operator/tensor/./../mxnet_op.h(701): here
            instantiation of "void mxnet::op::mxnet_op::mxnet_generic_kernel<OP,Args...>(int, Args...) [with OP=mxnet::op::mxnet_op::op_with_req<mxnet::op::mshadow_op::erfinv, 1>, Args=<uint8_t *, uint8_t *>]"
src/operator/tensor/./../mxnet_op.h(721): here
            instantiation of "void mxnet::op::mxnet_op::Kernel<OP, mxnet::gpu>::Launch(mshadow::Stream<mshadow::gpu> *, int, Args...) [with OP=mxnet::op::mxnet_op::op_with_req<mxnet::op::mshadow_op::erfinv, 1>, Args=<uint8_t *, uint8_t *>]"
src/operator/tensor/./elemwise_unary_op.h(243): here
            instantiation of "void mxnet::op::UnaryOp::Compute<xpu,OP>(const nnvm::NodeAttrs &, const mxnet::OpContext &, const std::vector<mxnet::TBlob, std::allocator<mxnet::TBlob>> &, const std::vector<mxnet::OpReqType, std::allocator<mxnet::OpReqType>> &, const std::vector<mxnet::TBlob, std::allocator<mxnet::TBlob>> &) [with xpu=mxnet::gpu, OP=mxnet::op::mshadow_op::erfinv]"
src/operator/tensor/elemwise_unary_op_basic.cu(67): here

) * add default behaviour for argmax * prototype of erfvin * add test * gpu support * Revert "add default behaviour for argmax" This reverts commit 64e9f1a. * move erfinv to contrib * edit copyright * remove atof * use std and update license * add license exclude file * fix per eric's comments * change license header

mseth10 · 2021-08-06T20:42:18Z

src/operator/contrib/erfinv-inl.h

+/*
+ * Copyright (c) 2014 Indiana University
+ * All rights reserved.
+ * Written by Prof. Gary L. Pavlis, Dept. of Geol. Sci.,
+ * Indiana University, Bloomington, IN
+ * This software is licensed under the New BSD license:
+ * Redistribution and use in source and binary forms,
+ * with or without modification, are permitted provided
+ * that the following conditions are met:
+ * Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the
+ * following disclaimer.
+ * Redistributions in binary form must reproduce the
+ * above copyright notice, this list of conditions and
+ * the following disclaimer in the documentation and/or
+ * other materials provided with the distribution.
+ * Neither the name of Indiana University nor
+ * the names of its contributors may be used to endorse
+ * or promote products derived from this software without
+ * specific prior written permission.
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
+ * CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
+ * PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
+ * THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY
+ * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
+ * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER
+ * IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
+ * USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+/*
+ * The next function is taken from
+ * https://github.com/antelopeusersgroup/antelope_contrib/blob/master/lib/location/libgenloc/erfinv.c.
+ * Output was modified to be inf or -inf when input is 1 or -1.
+ */


@rondogency do you remember where this license came from?

rondogency added 8 commits December 24, 2018 16:44

add default behaviour for argmax

64e9f1a

Merge branch 'master' of https://github.com/apache/incubator-mxnet

dc10b8d

prototype of erfvin

d610bfa

Merge branch 'master' of https://github.com/apache/incubator-mxnet

dc163c5

Merge branch 'master' into add-erfinv

09c2117

add test

57d2bba

gpu support

08aec73

Revert "add default behaviour for argmax"

006b2c8

This reverts commit 64e9f1a.

rondogency requested a review from szha as a code owner January 9, 2019 05:22

eric-haibin-lin reviewed Jan 9, 2019

View reviewed changes

src/operator/mshadow_op.h Outdated Show resolved Hide resolved

marcoabreu added the pr-awaiting-review PR is waiting for code review label Jan 9, 2019

eric-haibin-lin reviewed Jan 9, 2019

View reviewed changes

src/operator/mshadow_op.h Outdated Show resolved Hide resolved

move erfinv to contrib

5545e67

eric-haibin-lin reviewed Jan 13, 2019

View reviewed changes

src/operator/contrib/erfinv-inl.h Outdated Show resolved Hide resolved

rondogency added 3 commits January 13, 2019 18:22

edit copyright

722a573

remove atof

d878e07

use std and update license

cb3a217

eric-haibin-lin reviewed Jan 17, 2019

View reviewed changes

rondogency added 3 commits January 18, 2019 13:25

add license exclude file

9cc6c46

Merge branch 'master' of https://github.com/apache/incubator-mxnet

03357b9

resolve conflict in rat_excludes

2f2ff5c

eric-haibin-lin mentioned this pull request Jan 19, 2019

C++ test Core dump DROPOUT_PERF.TimingGPU #9857

Open

eric-haibin-lin reviewed Jan 19, 2019

View reviewed changes

src/operator/contrib/erfinv-inl.h Outdated Show resolved Hide resolved

src/operator/contrib/erfinv-inl.h Outdated Show resolved Hide resolved

fix per eric's comments

b012d84

eric-haibin-lin approved these changes Jan 21, 2019

View reviewed changes

tools/license_header.py Outdated Show resolved Hide resolved

change license header

647747b

eric-haibin-lin merged commit b86ccf1 into apache:master Jan 22, 2019

mseth10 reviewed Aug 6, 2021

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add erfinv operator for calculating inverse error function #13811

Add erfinv operator for calculating inverse error function #13811

rondogency commented Jan 9, 2019 •

edited

Loading

anirudhacharya commented Jan 9, 2019

eric-haibin-lin left a comment

rondogency commented Jan 19, 2019

eric-haibin-lin commented Jan 19, 2019

eric-haibin-lin left a comment

TaoLv commented Feb 21, 2019

mseth10 Aug 6, 2021

Add erfinv operator for calculating inverse error function #13811

Add erfinv operator for calculating inverse error function #13811

Conversation

rondogency commented Jan 9, 2019 • edited Loading

Description

Checklist

Essentials

Changes

Comments

anirudhacharya commented Jan 9, 2019

eric-haibin-lin left a comment

Choose a reason for hiding this comment

rondogency commented Jan 19, 2019

eric-haibin-lin commented Jan 19, 2019

eric-haibin-lin left a comment

Choose a reason for hiding this comment

TaoLv commented Feb 21, 2019

mseth10 Aug 6, 2021

Choose a reason for hiding this comment

rondogency commented Jan 9, 2019 •

edited

Loading