search

alashworth / test-issue-import

0 stars 0 forks source link

"call of overloaded ‘sqrt(int)’ is ambiguous" using g++ 7.2.0 #186

Open alashworth opened 5 years ago

alashworth commented 5 years ago

Issue by syclik Wednesday Feb 28, 2018 at 14:35 GMT Originally opened as https://github.com/stan-dev/stan/issues/2478

Summary:

A valid Stan program is failing to C++ compile on Linux using g++ 7.2.0.

This is failing from RStan, so we need to see if it's a Stan problem or an RStan problem.

(Originally reported by @zottelef on Discourse: http://discourse.mc-stan.org/t/qr-reparametrization-error/3426/9)

Description:

This program generates C++ code and then fails to compile.

data {
    int<lower=0> N;
    int<lower=0> K; //number of fixed effects
    matrix[N, K] x;
    vector[N] y;
}
transformed data {
    matrix[N, K] Q_ast;
    Q_ast = qr_Q(x)[, 1:K] * sqrt(N - 1);
}
parameters {
    real alpha; //intercept
    vector[K] theta; 
    real<lower=0> sigma;
}
model {
    y ~ normal(Q_ast * theta + alpha, sigma);
}

Reproducible Steps:

We need to determine if this is an RStan or a Stan problem. This error was reported using:

Current Output:

The relevant part of the compiler error is:

error: call of overloaded ‘sqrt(int)’ is ambiguous
             stan::math::assign(Q_ast, multiply(stan::model::rvalue(qr_Q(x), stan::model::cons_list(stan::model::index_omni(), stan::model::cons_list(stan::model::index_min_max(1, K), stan::model::nil_index_list())), "qr_Q(x)"),sqrt((N - 1))));

Here's the rest of the reported compiler error:

file25053f4922a5.cpp:130:240: error: call of overloaded ‘sqrt(int)’ is ambiguous
             stan::math::assign(Q_ast, multiply(stan::model::rvalue(qr_Q(x), stan::model::cons_list(stan::model::index_omni(), stan::model::cons_list(stan::model::index_min_max(1, K), stan::model::nil_index_list())), "qr_Q(x)"),sqrt((N - 1))));

...

Error in compileCode(f, code, language = language, verbose = verbose) : 
  Compilation ERROR, function(s)/method(s) not created! file25053f4922a5.cpp: In member function ‘void model250526cfdfd_climateTAVG_namespace::model250526cfdfd_climateTAVG::ctor_body(stan::io::var_context&, unsigned int, std::ostream*)’:
file25053f4922a5.cpp:130:240: error: call of overloaded ‘sqrt(int)’ is ambiguous
             stan::math::assign(Q_ast, multiply(stan::model::rvalue(qr_Q(x), stan::model::cons_list(stan::model::index_omni(), stan::model::cons_list(stan::model::index_min_max(1, K), stan::model::nil_index_list())), "qr_Q(x)"),sqrt((N - 1))));
                                                                                                                                                                                                                                                ^
In file included from /usr/include/features.h:419:0,
                 from /usr/include/x86_64-linux-gnu/bits/libc-header-start.h:33,
                 from /usr/include/stdint.h:
In addition: Warning message:
running command '/usr/lib/R/bin/R CMD SHLIB file25053f4922a5.cpp 2> file25053f4922a5.cpp.err.txt' had status 1

Expected Output:

No error.

Current Version:

v2.17.1

alashworth commented 5 years ago

Comment by bgoodri Wednesday Feb 28, 2018 at 15:15 GMT

It is a Stan (Math) issue and has already had several issues filed about it over the years, such as https://github.com/stan-dev/math/issues/712

alashworth commented 5 years ago

Comment by bob-carpenter Wednesday Feb 28, 2018 at 16:12 GMT

Closed. See #712 as @bgoodri pointed out.

I'll try to prioritize getting this fixed. Do you know if that compiler is available on Mac OSX or if there's a way to test with it easily?

alashworth commented 5 years ago

Comment by syclik Wednesday Feb 28, 2018 at 16:25 GMT

I spun up a docker image and tested it with Ubuntu with g++ 7.2.0. It works fine with CmdStan.

I'm trying to reproduce the error in RStan, but having trouble installing RStan under that configuration.

alashworth commented 5 years ago

Comment by zottelef Wednesday Feb 28, 2018 at 16:30 GMT

if you think I could be of any help (I am a really entry-level STAN user) let me know!

alashworth commented 5 years ago

Comment by syclik Wednesday Feb 28, 2018 at 16:49 GMT

@zottelef, thank you!

There is something you could try to help us figure this out. Can you:

  1. download the latest version of CmdStan (unpack this tarball: CmdStan 2.17.1)
  2. untar it
  3. go into that directory and write into a file called make/local two lines: 
     CXX=g++-7
 CC=g++-7
  4. type make build (or if you have many cores: make -j8 build)
  5. copy the Stan file to somewhere in the cmdstan directory, e.g. issue-2478.stan
  6. type make issue-2478
  7. report what happens when you type that.

What I'm trying to tease apart is whether this is a configuration problem in RStan or CmdStan / Stan.

alashworth commented 5 years ago

Comment by syclik Wednesday Feb 28, 2018 at 16:54 GMT

@bgoodri, this is the result I have from building from CmdStan (tagged version v2.17.1):

# make issue-2478 

--- Translating Stan model to C++ code ---
bin/stanc  issue-2478.stan --o=issue-2478.hpp
Model name=issue_2478_model
Input file=issue-2478.stan
Output file=issue-2478.hpp
Compiling pre-compiled header
g++-7 -Wall -I . -isystem stan/lib/stan_math/lib/eigen_3.3.3 -isystem stan/lib/stan_math/lib/boost_1.64.0 -isystem stan/lib/stan_math/lib/cvodes_2.9.0/include -std=c++1y -DBOOST_RESULT_OF_USE_TR1 -DBOOST_NO_DECLTYPE -DBOOST_DISABLE_ASSERTS -DBOOST_PHOENIX_NO_VARIADIC_EXPRESSION -Wno-unused-function -Wno-uninitialized -I src -isystem stan/src -isystem stan/lib/stan_math/ -DFUSION_MAX_VECTOR_SIZE=12 -Wno-unused-local-typedefs -DEIGEN_NO_DEBUG -DNO_FPRINTF_OUTPUT -pipe  -c -O3 stan/src/stan/model/model_header.hpp -o stan/src/stan/model/model_header.hpp.gch

--- Linking C++ model ---
g++-7 -Wall -I . -isystem stan/lib/stan_math/lib/eigen_3.3.3 -isystem stan/lib/stan_math/lib/boost_1.64.0 -isystem stan/lib/stan_math/lib/cvodes_2.9.0/include -std=c++1y -DBOOST_RESULT_OF_USE_TR1 -DBOOST_NO_DECLTYPE -DBOOST_DISABLE_ASSERTS -DBOOST_PHOENIX_NO_VARIADIC_EXPRESSION -Wno-unused-function -Wno-uninitialized -I src -isystem stan/src -isystem stan/lib/stan_math/ -DFUSION_MAX_VECTOR_SIZE=12 -Wno-unused-local-typedefs -DEIGEN_NO_DEBUG -DNO_FPRINTF_OUTPUT -pipe -lpthread   -O3 -o issue-2478 src/cmdstan/main.cpp -include issue-2478.hpp stan/lib/stan_math/lib/cvodes_2.9.0/lib/libsundials_nvecserial.a stan/lib/stan_math/lib/cvodes_2.9.0/lib/libsundials_cvodes.a

No compiler error. 

# g++-7 -v
Using built-in specs.
COLLECT_GCC=g++-7
COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-linux-gnu/7/lto-wrapper
OFFLOAD_TARGET_NAMES=nvptx-none
OFFLOAD_TARGET_DEFAULT=1
Target: x86_64-linux-gnu
Configured with: ../src/configure -v --with-pkgversion='Ubuntu 7.2.0-1ubuntu1~16.04' --with-bugurl=file:///usr/share/doc/gcc-7/README.Bugs --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --prefix=/usr --with-gcc-major-version-only --program-suffix=-7 --program-prefix=x86_64-linux-gnu- --enable-shared --enable-linker-build-id --libexecdir=/usr/lib --without-included-gettext --enable-threads=posix --libdir=/usr/lib --enable-nls --with-sysroot=/ --enable-clocale=gnu --enable-libstdcxx-debug --enable-libstdcxx-time=yes --with-default-libstdcxx-abi=new --enable-gnu-unique-object --disable-vtable-verify --enable-libmpx --enable-plugin --with-system-zlib --with-target-system-zlib --enable-objc-gc=auto --enable-multiarch --disable-werror --with-arch-32=i686 --with-abi=m64 --with-multilib-list=m32,m64,mx32 --enable-multilib --with-tune=generic --enable-offload-targets=nvptx-none --without-cuda-driver --enable-checking=release --build=x86_64-linux-gnu --host=x86_64-linux-gnu --target=x86_64-linux-gnu
Thread model: posix
gcc version 7.2.0 (Ubuntu 7.2.0-1ubuntu1~16.04)
alashworth commented 5 years ago

Comment by syclik Wednesday Feb 28, 2018 at 16:56 GMT

@zottelef, please let me know if you run into any trouble getting that up and running. It took me less than 30 minutes, but I know CmdStan inside and out.

alashworth commented 5 years ago

Comment by bob-carpenter Wednesday Feb 28, 2018 at 16:59 GMT

Thanks. Yikes. How in the world would R be messing things up? Is it something in the Makevars?

alashworth commented 5 years ago

Comment by bgoodri Wednesday Feb 28, 2018 at 19:16 GMT

goodrich@T540p:/opt/cmdstan$ make /tmp/sqrt

--- Translating Stan model to C++ code ---
bin/stanc  /tmp/sqrt.stan --o=/tmp/sqrt.hpp
Model name=sqrt_model
Input file=/tmp/sqrt.stan
Output file=/tmp/sqrt.hpp

--- Linking C++ model ---
g++-7 -Wall -I . -isystem stan/lib/stan_math/lib/eigen_3.3.3 -isystem stan/lib/stan_math/lib/boost_1.64.0 -isystem stan/lib/stan_math/lib/cvodes_2.9.0/include -std=c++1y -DBOOST_RESULT_OF_USE_TR1 -DBOOST_NO_DECLTYPE -DBOOST_DISABLE_ASSERTS -DBOOST_PHOENIX_NO_VARIADIC_EXPRESSION -I src -I stan/src -isystem stan/lib/stan_math/ -DEIGEN_NO_DEBUG -DFUSION_MAX_VECTOR_SIZE=12 -DNO_FPRINTF_OUTPUT -pipe -lpthread  -O3 -o /tmp/sqrt src/cmdstan/main.cpp -include /tmp/sqrt.hpp stan/lib/stan_math/lib/cvodes_2.9.0/lib/libsundials_nvecserial.a stan/lib/stan_math/lib/cvodes_2.9.0/lib/libsundials_cvodes.a
In file included from <command-line>:0:0:
/tmp/sqrt.hpp: In member function ‘void sqrt_model_namespace::sqrt_model::ctor_body(stan::io::var_context&, unsigned int, std::ostream*)’:
/tmp/sqrt.hpp:125:240: error: call of overloaded ‘sqrt(int)’ is ambiguous
 :model::rvalue(qr_Q(x), stan::model::cons_list(stan::model::index_omni(), stan::model::cons_list(stan::model::index_min_max(1, K), stan::model::nil_index_list())), "qr_Q(x)"),sqrt((N - 1))));
                                                                                                                                                                                            ^
In file included from /usr/include/features.h:419:0,
                 from /usr/include/x86_64-linux-gnu/bits/libc-header-start.h:33,
                 from /usr/include/stdint.h:26,
                 from /usr/lib/gcc/x86_64-linux-gnu/7/include/stdint.h:9,
                 from stan/lib/stan_math/stan/math/memory/stack_alloc.hpp:6,
                 from stan/lib/stan_math/stan/math/rev/core/autodiffstackstorage.hpp:4,
                 from stan/lib/stan_math/stan/math/rev/core.hpp:4,
                 from stan/lib/stan_math/stan/math/rev/mat.hpp:4,
                 from stan/lib/stan_math/stan/math.hpp:4,
                 from stan/src/stan/model/model_header.hpp:4,
                 from /tmp/sqrt.hpp:3,
                 from <command-line>:0:
/usr/include/x86_64-linux-gnu/bits/mathcalls.h:149:1: note: candidate: double sqrt(double)
 __MATHCALL (sqrt,, (_Mdouble_ __x));
 ^
In file included from stan/lib/stan_math/lib/boost_1.64.0/boost/config/no_tr1/cmath.hpp:21:0,
                 from stan/lib/stan_math/lib/boost_1.64.0/boost/math/tools/config.hpp:19,
                 from stan/lib/stan_math/stan/math/rev/core/var.hpp:7,
                 from stan/lib/stan_math/stan/math/rev/core/gevv_vvv_vari.hpp:5,
                 from stan/lib/stan_math/stan/math/rev/core.hpp:12,
                 from stan/lib/stan_math/stan/math/rev/mat.hpp:4,
                 from stan/lib/stan_math/stan/math.hpp:4,
                 from stan/src/stan/model/model_header.hpp:4,
                 from /tmp/sqrt.hpp:3,
                 from <command-line>:0:
/usr/include/c++/7/cmath:463:3: note: candidate: constexpr float std::sqrt(float)
   sqrt(float __x)
   ^~~~
/usr/include/c++/7/cmath:467:3: note: candidate: constexpr long double std::sqrt(long double)
   sqrt(long double __x)
   ^~~~
/usr/include/c++/7/cmath:475:5: note: candidate: constexpr typename __gnu_cxx::__enable_if<std::__is_integer<_Tp>::__value, double>::__type std::sqrt(_Tp) [with _Tp = int; typename __gnu_cxx::__enable_if<std::__is_integer<_Tp>::__value, double>::__type = double]
     sqrt(_Tp __x)
     ^~~~
In file included from stan/lib/stan_math/stan/math/rev/scal/fun/pow.hpp:5:0,
                 from stan/lib/stan_math/stan/math/rev/scal/fun/inc_beta.hpp:7,
                 from stan/lib/stan_math/stan/math/rev/scal/fun/grad_inc_beta.hpp:12,
                 from stan/lib/stan_math/stan/math/rev/scal.hpp:50,
                 from stan/lib/stan_math/stan/math/rev/arr.hpp:9,
                 from stan/lib/stan_math/stan/math/rev/mat.hpp:13,
                 from stan/lib/stan_math/stan/math.hpp:4,
                 from stan/src/stan/model/model_header.hpp:4,
                 from /tmp/sqrt.hpp:3,
                 from <command-line>:0:
stan/lib/stan_math/stan/math/rev/scal/fun/sqrt.hpp:50:16: note: candidate: stan::math::var stan::math::sqrt(const stan::math::var&)
     inline var sqrt(const var& a) {
                ^~~~
In file included from stan/lib/stan_math/stan/math/prim/mat.hpp:209:0,
                 from stan/lib/stan_math/stan/math/rev/mat.hpp:12,
                 from stan/lib/stan_math/stan/math.hpp:4,
                 from stan/src/stan/model/model_header.hpp:4,
                 from /tmp/sqrt.hpp:3,
                 from <command-line>:0:
stan/lib/stan_math/stan/math/prim/mat/fun/sqrt.hpp:32:5: note: candidate: typename stan::math::apply_scalar_unary<stan::math::sqrt_fun, T>::return_t stan::math::sqrt(const T&) [with T = int; typename stan::math::apply_scalar_unary<stan::math::sqrt_fun, T>::return_t = double]
     sqrt(const T& x) {
     ^~~~
make: *** [make/models:12: /tmp/sqrt] Error 1
alashworth commented 5 years ago

Comment by bob-carpenter Wednesday Feb 28, 2018 at 20:18 GMT

The solution's the same here. Write our own sqrt() and make sure we call that.

On Feb 28, 2018, at 2:16 PM, bgoodri notifications@github.com wrote:

goodrich@T540p:/opt/cmdstan$ make /tmp/sqrt

--- Translating Stan model to C++ code --- bin/stanc /tmp/sqrt.stan --o=/tmp/sqrt.hpp Model name=sqrt_model Input file=/tmp/sqrt.stan Output file=/tmp/sqrt.hpp

--- Linking C++ model --- g++-7 -Wall -I . -isystem stan/lib/stan_math/lib/eigen_3.3.3 -isystem stan/lib/stan_math/lib/boost_1.64.0 -isystem stan/lib/stan_math/lib/cvodes_2.9.0/include -std=c++1y -DBOOST_RESULT_OF_USE_TR1 -DBOOST_NO_DECLTYPE -DBOOST_DISABLE_ASSERTS -DBOOST_PHOENIX_NO_VARIADIC_EXPRESSION -I src -I stan/src -isystem stan/lib/stan_math/ -DEIGEN_NO_DEBUG -DFUSION_MAX_VECTOR_SIZE=12 -DNO_FPRINTF_OUTPUT -pipe -lpthread -O3 -o /tmp/sqrt src/cmdstan/main.cpp -include /tmp/sqrt.hpp stan/lib/stan_math/lib/cvodes_2.9.0/lib/libsundials_nvecserial.a stan/lib/stan_math/lib/cvodes_2.9.0/lib/libsundials_cvodes.a In file included from :0:0: /tmp/sqrt.hpp: In member function ‘void sqrt_model_namespace::sqrt_model::ctor_body(stan::io::var_context&, unsigned int, std::ostream*)’: /tmp/sqrt.hpp:125:240: error: call of overloaded ‘sqrt(int)’ is ambiguous :model::rvalue(qr_Q(x), stan::model::cons_list(stan::model::index_omni(), stan::model::cons_list(stan::model::index_min_max(1, K), stan::model::nil_index_list())), "qr_Q(x)"),sqrt((N - 1)))); ^ In file included from /usr/include/features.h:419:0, from /usr/include/x86_64-linux-gnu/bits/libc-header-start.h:33, from /usr/include/stdint.h:26, from /usr/lib/gcc/x86_64-linux-gnu/7/include/stdint.h:9, from stan/lib/stan_math/stan/math/memory/stack_alloc.hpp:6, from stan/lib/stan_math/stan/math/rev/core/autodiffstackstorage.hpp:4, from stan/lib/stan_math/stan/math/rev/core.hpp:4, from stan/lib/stan_math/stan/math/rev/mat.hpp:4, from stan/lib/stan_math/stan/math.hpp:4, from stan/src/stan/model/model_header.hpp:4, from /tmp/sqrt.hpp:3, from :0: /usr/include/x86_64-linux-gnu/bits/mathcalls.h:149:1: note: candidate: double sqrt(double) MATHCALL (sqrt,, (Mdouble x)); ^ In file included from stan/lib/stan_math/lib/boost_1.64.0/boost/config/no_tr1/cmath.hpp:21:0, from stan/lib/stan_math/lib/boost_1.64.0/boost/math/tools/config.hpp:19, from stan/lib/stan_math/stan/math/rev/core/var.hpp:7, from stan/lib/stan_math/stan/math/rev/core/gevv_vvv_vari.hpp:5, from stan/lib/stan_math/stan/math/rev/core.hpp:12, from stan/lib/stan_math/stan/math/rev/mat.hpp:4, from stan/lib/stan_math/stan/math.hpp:4, from stan/src/stan/model/model_header.hpp:4, from /tmp/sqrt.hpp:3, from :0: /usr/include/c++/7/cmath:463:3: note: candidate: constexpr float std::sqrt(float) sqrt(float x) ^~~~ /usr/include/c++/7/cmath:467:3: note: candidate: constexpr long double std::sqrt(long double) sqrt(long double x) ^~~~ /usr/include/c++/7/cmath:475:5: note: candidate: constexpr typename gnu_cxx::__enable_if<std::is_integer<_Tp>::value, double>::type std::sqrt(_Tp) [with _Tp = int; typename gnu_cxx::__enable_if<std::is_integer<_Tp>::value, double>::type = double] sqrt(_Tp __x) ^~~~ In file included from stan/lib/stan_math/stan/math/rev/scal/fun/pow.hpp:5:0, from stan/lib/stan_math/stan/math/rev/scal/fun/inc_beta.hpp:7, from stan/lib/stan_math/stan/math/rev/scal/fun/grad_inc_beta.hpp:12, from stan/lib/stan_math/stan/math/rev/scal.hpp:50, from stan/lib/stan_math/stan/math/rev/arr.hpp:9, from stan/lib/stan_math/stan/math/rev/mat.hpp:13, from stan/lib/stan_math/stan/math.hpp:4, from stan/src/stan/model/model_header.hpp:4, from /tmp/sqrt.hpp:3, from :0: stan/lib/stan_math/stan/math/rev/scal/fun/sqrt.hpp:50:16: note: candidate: stan::math::var stan::math::sqrt(const stan::math::var&) inline var sqrt(const var& a) { ^~~~ In file included from stan/lib/stan_math/stan/math/prim/mat.hpp:209:0, from stan/lib/stan_math/stan/math/rev/mat.hpp:12, from stan/lib/stan_math/stan/math.hpp:4, from stan/src/stan/model/model_header.hpp:4, from /tmp/sqrt.hpp:3, from :0: stan/lib/stan_math/stan/math/prim/mat/fun/sqrt.hpp:32:5: note: candidate: typename stan::math::apply_scalar_unary<stan::math::sqrt_fun, T>::return_t stan::math::sqrt(const T&) [with T = int; typename stan::math::apply_scalar_unary<stan::math::sqrt_fun, T>::return_t = double] sqrt(const T& x) { ^~~~ make: *** [make/models:12: /tmp/sqrt] Error 1

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub, or mute the thread.

alashworth commented 5 years ago

Comment by syclik Wednesday Feb 28, 2018 at 21:02 GMT

@bgoodri, is that compiler installed on the linux box?

If so, I think we can run a test in Stan (compile all models) with Jenkins that will test that things are C++ compilable with that compiler.

alashworth commented 5 years ago

Comment by syclik Wednesday Feb 28, 2018 at 21:02 GMT

(I can't reproduce the failure with g++ 7.2.0.)

alashworth commented 5 years ago

Comment by bgoodri Wednesday Feb 28, 2018 at 21:06 GMT

Yes

On Wed, Feb 28, 2018 at 4:02 PM, Daniel Lee notifications@github.com wrote:

@bgoodri https://github.com/bgoodri, is that compiler installed on the linux box?

If so, I think we can run a test in Stan (compile all models) with Jenkins that will test that things are C++ compilable with that compiler.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/stan-dev/stan/issues/2478#issuecomment-369381498, or mute the thread https://github.com/notifications/unsubscribe-auth/ADOrqiVqlorn9SiB9foEJPc_h1wxV-ydks5tZb7QgaJpZM4SWwLs .

alashworth commented 5 years ago

Comment by syclik Thursday Mar 01, 2018 at 19:56 GMT

@bgoodri, mind doing me a favor and seeing if this test fails in Stan? (after setting CXX = g++-7 in make/local)

./runTests.py src/test/integration

We've written out a lot of Stan programs that we assume to be compilable by the C++ compiler and the two tests in there make sure they're parsed properly and that they compile. If that catches when it doesn't work, we can put that into a Jenkins job and make sure we are able to keep compiling with changes to the code base.

alashworth commented 5 years ago

Comment by zottelef Friday Mar 02, 2018 at 08:39 GMT

@syclik : sorry for the late answer. I did not notice the advances in the topic. Really sorry about that. However, following the work flow you gave me I obtain:

> make issue-2478

--- Translating Stan model to C++ code ---
bin/stanc  issue-2478.stan --o=issue-2478.hpp
Model name=issue_2478_model
Input file=issue-2478.stan
Output file=issue-2478.hpp
Compiling pre-compiled header
g++-7 -Wall -I . -isystem stan/lib/stan_math/lib/eigen_3.3.3 -isystem stan/lib/stan_math/lib/boost_1.64.0 -isystem stan/lib/stan_math/lib/cvodes_2.9.0/include -std=c++1y -DBOOST_RESULT_OF_USE_TR1 -DBOOST_NO_DECLTYPE -DBOOST_DISABLE_ASSERTS -DBOOST_PHOENIX_NO_VARIADIC_EXPRESSION -Wno-unused-function -Wno-uninitialized -I src -isystem stan/src -isystem stan/lib/stan_math/ -DFUSION_MAX_VECTOR_SIZE=12 -Wno-unused-local-typedefs -DEIGEN_NO_DEBUG -DNO_FPRINTF_OUTPUT -pipe  -c -O3 stan/src/stan/model/model_header.hpp -o stan/src/stan/model/model_header.hpp.gch

--- Linking C++ model ---
g++-7 -Wall -I . -isystem stan/lib/stan_math/lib/eigen_3.3.3 -isystem stan/lib/stan_math/lib/boost_1.64.0 -isystem stan/lib/stan_math/lib/cvodes_2.9.0/include -std=c++1y -DBOOST_RESULT_OF_USE_TR1 -DBOOST_NO_DECLTYPE -DBOOST_DISABLE_ASSERTS -DBOOST_PHOENIX_NO_VARIADIC_EXPRESSION -Wno-unused-function -Wno-uninitialized -I src -isystem stan/src -isystem stan/lib/stan_math/ -DFUSION_MAX_VECTOR_SIZE=12 -Wno-unused-local-typedefs -DEIGEN_NO_DEBUG -DNO_FPRINTF_OUTPUT -pipe -lpthread   -O3 -o issue-2478 src/cmdstan/main.cpp -include issue-2478.hpp stan/lib/stan_math/lib/cvodes_2.9.0/lib/libsundials_nvecserial.a stan/lib/stan_math/lib/cvodes_2.9.0/lib/libsundials_cvodes.a
alashworth commented 5 years ago

Comment by syclik Friday Mar 02, 2018 at 12:25 GMT

Thank you!

alashworth commented 5 years ago

Comment by bob-carpenter Wednesday Mar 14, 2018 at 07:08 GMT

OK, this is the one I meant to reopen, as the problem's in this repo.

alashworth commented 5 years ago

Comment by stenknutsen Monday Oct 15, 2018 at 14:48 GMT

Summary: Compilation fails when using sqrt().

Description: After compilation failed, I replaced sqrt() with pow(x , 0.5) and everything ran fine.

Reproducible Steps: Tried running Gaussian_Stan.R from https://github.com/stan-dev/example-models/tree/master/Bayesian_Cognitive_Modeling/ParameterEstimation/Gaussian

# clears workspace: 
rm(list=ls()) 

library(rstan)
Sys.setenv(USE_CXX14=1)
#### Notes to Stan model #######################################################
## 1) If parameter's prior distribution is not specified, Stan will assume that
##    you want it to be distributed uniformly with boundaries given by variable 
##    constraints. Here constrains <lower=0,upper=10> give uniform (0, 100)
## 2) In Stan, most of the sampling statements can be vectorized. In this example
##    you can see it in the statement for vector x. Instead of using for loop for  
##    each element of the vector, we can simple write it as above. This saves code, 
##    speeds up computation. For more information read Vectorization chapter in
##    the Stan manual (p.231 in version 2.4.0)
################################################################################
model <- "
// Inferring the Mean and Standard Deviation of a Gaussian
data { 
  int<lower=1> n;
  vector<lower=0>[n] x;
}
parameters {
  real mu;
  real<lower=0,upper=10> sigma; 
} 
model {
  // Priors
  mu ~ normal(0, sqrt(1000)); #### Looks like it gets confused by sqrt(1000) here -- pow(1000 , 0.5) works though.

  // Data Come From A Gaussian
  x ~ normal(mu, sigma);
}"

x <- c(1.1, 1.9, 2.3, 1.8)
n <- length(x)

data <- list(x=x, n=n) # to be passed on to Stan
myinits <- list(
  list(mu=0, sigma=1))

# parameters to be monitored: 
parameters <- c("mu", "sigma")

# The following command calls Stan with specific options.
# For a detailed description type "?rstan".
samples <- stan(model_code=model,   
                data=data, 
                init=myinits,  # If not specified, gives random inits
                pars=parameters,
                iter=10000, 
                chains=1, 
                thin=1,
                # warmup = 100,  # Stands for burn-in; Default = iter/2
                # seed = 123  # Setting seed; Default is random seed
                )
# Now the values for the monitored parameters are in the "samples" object, 
# ready for inspection.

mu <- extract(samples)$mu
sigma <- extract(samples)$sigma

Current Output: Relevant error messages:

In file included from /Library/Frameworks/R.framework/Versions/3.5/Resources/library/StanHeaders/include/stan/math/prim/mat/fun/csr_extract_u.hpp:6:
In file included from /Library/Frameworks/R.framework/Versions/3.5/Resources/library/RcppEigen/include/Eigen/Sparse:33:
In file included from /Library/Frameworks/R.framework/Versions/3.5/Resources/library/RcppEigen/include/Eigen/IterativeLinearSolvers:46:
/Library/Frameworks/R.framework/Versions/3.5/Resources/library/RcppEigen/include/Eigen/src/Core/util/ReenableStupidWarnings.h:10:30: warning: pragma diagnostic pop could not pop, no matching push [-Wunknown-pragmas]
    #pragma clang diagnostic pop
                             ^
file2de046558108.cpp:211:56: error: call to 'sqrt' is ambiguous
            lp_accum__.add(normal_log<propto__>(mu, 0, sqrt(1000)));
                                                       ^~~~
/usr/local/clang4/bin/../include/c++/v1/math.h:917:1: note: candidate function [with _A1 = int]
sqrt(_A1 __lcpp_x) _NOEXCEPT {return sqrt((double)__lcpp_x);}
^
/Library/Frameworks/R.framework/Versions/3.5/Resources/library/StanHeaders/include/stan/math/prim/mat/fun/sqrt.hpp:31:59: note: candidate function [with T = int]
inline typename apply_scalar_unary<sqrt_fun, T>::return_t sqrt(const T& x) {
                                                          ^
/usr/include/math.h:447:15: note: candidate function
extern double sqrt(double);
              ^
/usr/local/clang4/bin/../include/c++/v1/math.h:910:46: note: candidate function
inline _LIBCPP_INLINE_VISIBILITY float       sqrt(float __lcpp_x) _NOEXCEPT       {return sqrtf(__lcpp_x);}
                                             ^
/usr/local/clang4/bin/../include/c++/v1/math.h:911:46: note: candidate function
inline _LIBCPP_INLINE_VISIBILITY long double sqrt(long double __lcpp_x) _NOEXCEPT {return sqrtl(__lcpp_x);}
                                             ^
/Library/Frameworks/R.framework/Versions/3.5/Resources/library/StanHeaders/include/stan/math/rev/scal/fun/sqrt.hpp:46:12: note: candidate function
inline var sqrt(const var& a) { return var(new sqrt_vari(a.vi_)); }
           ^
13 warnings and 1 error generated.
make: *** [file2de046558108.o] Error 1

ERROR(s) during compilation: source code errors or compiler configuration errors!

and

Error in compileCode(f, code, language = language, verbose = verbose) : Compilation ERROR, function(s)/method(s) not created! In file included from file2de046558108.cpp:8: In file included from /Library/Frameworks/R.framework/Versions/3.5/Resources/library/StanHeaders/include/src/stan/model/model_header.hpp:4: In file included from /Library/Frameworks/R.framework/Versions/3.5/Resources/library/StanHeaders/include/stan/math.hpp:4: In file included from /Library/Frameworks/R.framework/Versions/3.5/Resources/library/StanHeaders/include/stan/math/rev/mat.hpp:4: In file included from /Library/Frameworks/R.framework/Versions/3.5/Resources/library/StanHeaders/include/stan/math/rev/core.hpp:14: In file included from /Library/Frameworks/R.framework/Versions/3.5/Resources/library/StanHeaders/include/stan/math/rev/core/matrix_vari.hpp:4: In file included from /Library/Frameworks/R.framework/Versions/3.5/Resources/library/StanHeaders/include/stan/math/rev/mat/fun/Eigen_NumTraits.hpp:4: In file included from /Library/Frameworks/R.framework/Versions/3.5/Resour

Expected Output: Two arrays populated with values: mu and sigma

Current Version: v2.17.4