Compiling error after upgrading to MacOS Big Sur

Interfaces PyStan
1

Hello,

I’ve been using PySTAN for over a year. However, I started getting compilation errors right after I performed Apple’s software update (upgrade to Big Sur):

command 'x86_64-apple-darwin13.4.0-clang++' failed with exit status 1

I tried many solution suggestions such as uninstalling and reinstalling anaconda, followed by:

conda create -n stan_env python=3.7 -c conda-forge
conda activate stan_env
conda install pystan -c conda-forge

conda install clang_osx-64 clangxx_osx-64 -c conda-forge
export CC=x86_64-apple-darwin13.4.0-clang
export CXX=x86_64-apple-darwin13.4.0-clang++

but to no avail, just changing the exit status to 1 which was something like 240 before…

I’m using this setup:

  • MacOS Big Sur 11.0.1 (I was using MacOS Catalina 10.15.7 before the upgrade)
  • Python 3.7.6
  • PyStan 2.19.1.1

Thank you in advance.

1 Like
2

Thanks for the info.

Can you first try pip install pystan in that env?

(CmdStanR has been working normally --> CmdStanPy should work)

This might be some error with compiler or python+compiler.

If you use verbose=True, do you see any C++ errors (see terminal if you use jupyter)

3

I first tried pip install pystan as you suggested, but the resulting error message is the same.

Even when I use verbose=False, there are so many lines flowing on the screen while compiling which I have never seen before even though as I said before I’m rigorously using pystan for more than a year. This is the last few lines when I switch to verbose=True:

152 warnings generated. x86_64-apple-darwin13.4.0-clang++ -bundle -undefined dynamic_lookup -Wl,-pie -Wl,-headerpad_max_install_names -Wl,-dead_strip_dylibs -Wl,-rpath,/opt/anaconda3/envs/stan_env/lib -L/opt/anaconda3/envs/stan_env/lib -flto -Wl,-export_dynamic -Wl,-pie -Wl,-headerpad_max_install_names -Wl,-dead_strip_dylibs -Wl,-rpath,/opt/anaconda3/envs/stan_env/lib -L/opt/anaconda3/envs/stan_env/lib -Wl,-pie -Wl,-headerpad_max_install_names -Wl,-dead_strip_dylibs -Wl,-rpath,/opt/anaconda3/envs/stan_env/lib -L/opt/anaconda3/envs/stan_env/lib -march=core2 -mtune=haswell -mssse3 -ftree-vectorize -fPIC -fPIE -fstack-protector-strong -O2 -pipe -isystem /opt/anaconda3/envs/stan_env/include -D_FORTIFY_SOURCE=2 -mmacosx-version-min=10.9 -isystem /opt/anaconda3/envs/stan_env/include -arch x86_64 /var/folders/5h/05bld9s94p9f2yts8__07jtwp00nfd/T/pystan__h6c4o3z/var/folders/5h/05bld9s94p9f2yts8__07jtwp00nfd/T/pystan__h6c4o3z/stanfit4anon_model_5944b02c79788fa0db5b3a93728ca2bf_7451425786530191539.o -o /var/folders/5h/05bld9s94p9f2yts8__07jtwp00nfd/T/pystan__h6c4o3z/stanfit4anon_model_5944b02c79788fa0db5b3a93728ca2bf_7451425786530191539.cpython-37m-darwin.so ld: warning: -pie being ignored. It is only used when linking a main executable ld: unsupported tapi file type '!tapi-tbd' in YAML file '/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/usr/lib/libSystem.tbd' for architecture x86_64 clang-10: error: linker command failed with exit code 1 (use -v to see invocation)

That clang-10:error is the only thing I saw with ‘error’ in it, so I’m not sure I’m seeing any C++ errors. I skimmed through the output in the terminal and it’s full of things like this:

/opt/anaconda3/envs/stan_env/lib/python3.7/site-packages/pystan/stan/src/stan/services/experimental/advi/meanfield.hpp:87:20: note: in instantiation of member function 'stan::variational::advi<anon_model_5944b02c79788fa0db5b3a93728ca2bf_namespace::anon_model_5944b02c79788fa0db5b3a93728ca2bf, stan::variational::normal_meanfield, boost::random::additive_combine_engine<boost::random::linear_congruential_engine<unsigned int, 40014, 0, 2147483563>, boost::random::linear_congruential_engine<unsigned int, 40692, 0, 2147483399> > >::run' requested here cmd_advi.run(eta, adapt_engaged, adapt_iterations, ^ /opt/anaconda3/envs/stan_env/lib/python3.7/site-packages/pystan/stan_fit.hpp:1388:15: note: in instantiation of function template specialization 'stan::services::experimental::advi::meanfield<anon_model_5944b02c79788fa0db5b3a93728ca2bf_namespace::anon_model_5944b02c79788fa0db5b3a93728ca2bf>' requested here ::meanfield(model, *init_context_ptr, ^ /opt/anaconda3/envs/stan_env/lib/python3.7/site-packages/pystan/stan_fit.hpp:1649:15: note: in instantiation of function template specialization 'pystan::(anonymous namespace)::command<anon_model_5944b02c79788fa0db5b3a93728ca2bf_namespace::anon_model_5944b02c79788fa0db5b3a93728ca2bf, boost::random::additive_combine_engine<boost::random::linear_congruential_engine<unsigned int, 40014, 0, 2147483563>, boost::random::linear_congruential_engine<unsigned int, 40692, 0, 2147483399> > >' requested here ret = command(args, model_, holder, names_oi_tidx_, ^ /var/folders/5h/05bld9s94p9f2yts8__07jtwp00nfd/T/pystan__h6c4o3z/stanfit4anon_model_5944b02c79788fa0db5b3a93728ca2bf_7451425786530191539.cpp:9486:34: note: in instantiation of member function 'pystan::stan_fit<anon_model_5944b02c79788fa0db5b3a93728ca2bf_namespace::anon_model_5944b02c79788fa0db5b3a93728ca2bf, boost::random::additive_combine_engine<boost::random::linear_congruential_engine<unsigned int, 40014, 0, 2147483563>, boost::random::linear_congruential_engine<unsigned int, 40692, 0, 2147483399> > >::call_sampler' requested here __pyx_t_15 = __pyx_v_fitptr->call_sampler((*__pyx_v_argsptr), (*__pyx_v_holderptr)); ^ In file included from /var/folders/5h/05bld9s94p9f2yts8__07jtwp00nfd/T/pystan__h6c4o3z/stanfit4anon_model_5944b02c79788fa0db5b3a93728ca2bf_7451425786530191539.cpp:702: In file included from /opt/anaconda3/envs/stan_env/lib/python3.7/site-packages/pystan/py_var_context.hpp:12: In file included from /opt/anaconda3/envs/stan_env/lib/python3.7/site-packages/pystan/stan/src/stan/io/dump.hpp:6: In file included from /opt/anaconda3/envs/stan_env/lib/python3.7/site-packages/pystan/stan/lib/stan_math/stan/math/prim/mat.hpp:336: In file included from /opt/anaconda3/envs/stan_env/lib/python3.7/site-packages/pystan/stan/lib/stan_math/stan/math/prim/mat/prob/poisson_log_glm_log.hpp:5: /opt/anaconda3/envs/stan_env/lib/python3.7/site-packages/pystan/stan/lib/stan_math/stan/math/prim/mat/prob/poisson_log_glm_lpmf.hpp:64:59: warning: unused typedef 'T_alpha_val' [-Wunused-local-typedef] typename partials_return_type<T_alpha>::type>::type T_alpha_val;

Again, these are things that I’ve never encountered in Stan before.

4

Ok, something wrong with the xcode commandline tools.

I’m not sure if there are any instructions for Big Sur, but maybe this could help node.js - ld: unsupported tapi file type '!tapi-tbd' in YAML file - Stack Overflow

2 Likes
5

Spot on diagnosis! Thank you! Doing the following as it’s suggested in the link you sent:

sudo rm -rf /Library/Developer/CommandLineTools
xcode-select --install

then a clean installation of Anaconda and PyStan solved the issue.

Thank you so much!

3 Likes
8

Hi Mustafa, I did all of this, any other suggestion? I still get the problem while running a Stan model. I’m using sm = bebi103.stan.StanModel(model_code=bssm) for the model. bass is a Stan code like “”" data{} parameters{} … “”". What version of CommandLineTools are you using? I couldn’t downgrade to 11.5 as is suggested in node.js - ld: unsupported tapi file type '!tapi-tbd' in YAML file - Stack Overflow

1 Like
9

I just solved it, I followed your instructions again, this time uninstalling and installing again Anaconda.