pymc
pymc copied to clipboard
pymc-devs
Stochastic Volatility model
When I am doing the Stochastic Volatility Model example, I had a bug as follows:
It was at the step of executing the code of building the stochastic volatility model
stochastic_vol_model = make_stochastic_volatility_model(returns)
We'll need more info to help you. Which operating system? How did you install PyMC? Which version? Can you import pymc and run a simple model Form the Quickstart tutorial? What is the full error traceback?
I use windows and use anaconda to install PyMC3 I could import pymc but there is an error when defining the model
ImportError Traceback (most recent call last)
Input In [4], in <cell line: 3>()
1 N = len(hits)
3 with pm.Model() as baseball_model:
----> 5 phi = pm.Uniform("phi", lower=0.0, upper=1.0)
7 kappa_log = pm.Exponential("kappa_log", lam=1.5)
8 kappa = pm.Deterministic("kappa", tt.exp(kappa_log))
File ~\anaconda3\envs\pymc_env\lib\site-packages\pymc3\distributions\distribution.py:124, in Distribution.__new__(cls, name, *args, **kwargs)
122 dist = cls.dist(*args, **kwargs, shape=shape)
123 else:
--> 124 dist = cls.dist(*args, **kwargs)
125 return model.Var(name, dist, data, total_size, dims=dims)
File ~\anaconda3\envs\pymc_env\lib\site-packages\pymc3\distributions\distribution.py:133, in Distribution.dist(cls, *args, **kwargs)
130 @classmethod
131 def dist(cls, *args, **kwargs):
132 dist = object.__new__(cls)
--> 133 dist.__init__(*args, **kwargs)
134 return dist
File ~\anaconda3\envs\pymc_env\lib\site-packages\pymc3\distributions\continuous.py:228, in Uniform.__init__(self, lower, upper, *args, **kwargs)
226 self.lower = lower = tt.as_tensor_variable(floatX(lower))
227 self.upper = upper = tt.as_tensor_variable(floatX(upper))
--> 228 self.mean = (upper + lower) / 2.0
229 self.median = self.mean
231 super().__init__(lower=lower, upper=upper, *args, **kwargs)
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\tensor\var.py:101, in _tensor_py_operators.__add__(self, other)
99 def __add__(self, other):
100 try:
--> 101 return theano.tensor.basic.add(self, other)
102 # We should catch the minimum number of exception here.
103 # Otherwise this will convert error when Theano flags
104 # compute_test_value is used
105 # Evidently, we need to catch NotImplementedError
106 # TypeError from as_tensor_variable are caught in Elemwise.make_node
107 # Oterwise TensorVariable * SparseVariable won't work!
108 except (NotImplementedError, TypeError):
109 # We must return NotImplemented and not an
110 # NotImplementedError or raise an NotImplementedError.
111 # That way python will give a good error message like this
112 # `TypeError: unsupported operand type(s) for +:
113 # 'TensorVariable' and 'TensorVariable'`
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\graph\op.py:253, in Op.__call__(self, *inputs, **kwargs)
250 node = self.make_node(*inputs, **kwargs)
252 if config.compute_test_value != "off":
--> 253 compute_test_value(node)
255 if self.default_output is not None:
256 rval = node.outputs[self.default_output]
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\graph\op.py:126, in compute_test_value(node)
123 compute_map[o] = [False]
125 # Create a thunk that performs the computation
--> 126 thunk = node.op.make_thunk(node, storage_map, compute_map, no_recycling=[])
127 thunk.inputs = [storage_map[v] for v in node.inputs]
128 thunk.outputs = [storage_map[v] for v in node.outputs]
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\graph\op.py:634, in COp.make_thunk(self, node, storage_map, compute_map, no_recycling, impl)
630 self.prepare_node(
631 node, storage_map=storage_map, compute_map=compute_map, impl="c"
632 )
633 try:
--> 634 return self.make_c_thunk(node, storage_map, compute_map, no_recycling)
635 except (NotImplementedError, MethodNotDefined):
636 # We requested the c code, so don't catch the error.
637 if impl == "c":
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\graph\op.py:600, in COp.make_c_thunk(self, node, storage_map, compute_map, no_recycling)
598 print(f"Disabling C code for {self} due to unsupported float16")
599 raise NotImplementedError("float16")
--> 600 outputs = cl.make_thunk(
601 input_storage=node_input_storage, output_storage=node_output_storage
602 )
603 thunk, node_input_filters, node_output_filters = outputs
605 def rval():
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\link\c\basic.py:1203, in CLinker.make_thunk(self, input_storage, output_storage, storage_map)
1175 """
1176 Compiles this linker's fgraph and returns a function to perform the
1177 computations, as well as lists of storage cells for both the inputs
(...)
1200 first_output = ostor[0].data
1201 """
1202 init_tasks, tasks = self.get_init_tasks()
-> 1203 cthunk, module, in_storage, out_storage, error_storage = self.__compile__(
1204 input_storage, output_storage, storage_map
1205 )
1207 res = _CThunk(cthunk, init_tasks, tasks, error_storage, module)
1208 res.nodes = self.node_order
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\link\c\basic.py:1138, in CLinker.__compile__(self, input_storage, output_storage, storage_map)
1136 input_storage = tuple(input_storage)
1137 output_storage = tuple(output_storage)
-> 1138 thunk, module = self.cthunk_factory(
1139 error_storage,
1140 input_storage,
1141 output_storage,
1142 storage_map,
1143 )
1144 return (
1145 thunk,
1146 module,
(...)
1155 error_storage,
1156 )
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\link\c\basic.py:1634, in CLinker.cthunk_factory(self, error_storage, in_storage, out_storage, storage_map)
1632 for node in self.node_order:
1633 node.op.prepare_node(node, storage_map, None, "c")
-> 1634 module = get_module_cache().module_from_key(key=key, lnk=self)
1636 vars = self.inputs + self.outputs + self.orphans
1637 # List of indices that should be ignored when passing the arguments
1638 # (basically, everything that the previous call to uniq eliminated)
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\link\c\cmodule.py:1191, in ModuleCache.module_from_key(self, key, lnk)
1189 try:
1190 location = dlimport_workdir(self.dirname)
-> 1191 module = lnk.compile_cmodule(location)
1192 name = module.__file__
1193 assert name.startswith(location)
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\link\c\basic.py:1543, in CLinker.compile_cmodule(self, location)
1541 try:
1542 _logger.debug(f"LOCATION {location}")
-> 1543 module = c_compiler.compile_str(
1544 module_name=mod.code_hash,
1545 src_code=src_code,
1546 location=location,
1547 include_dirs=self.header_dirs(),
1548 lib_dirs=self.lib_dirs(),
1549 libs=libs,
1550 preargs=preargs,
1551 )
1552 except Exception as e:
1553 e.args += (str(self.fgraph),)
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\link\c\cmodule.py:2557, in GCC_compiler.compile_str(module_name, src_code, location, include_dirs, lib_dirs, libs, preargs, py_module, hide_symbols)
2555 open(os.path.join(location, "__init__.py"), "w").close()
2556 assert os.path.isfile(lib_filename)
-> 2557 return dlimport(lib_filename)
File ~\anaconda3\envs\pymc_env\lib\site-packages\theano\link\c\cmodule.py:294, in dlimport(fullpath, suffix)
292 with warnings.catch_warnings():
293 warnings.filterwarnings("ignore", message="numpy.ndarray size changed")
--> 294 rval = __import__(module_name, {}, {}, [module_name])
295 t1 = time.time()
296 import_time += t1 - t0
ImportError: DLL load failed while importing mf6917bb35eaa79d4a20c20b9dc13c0435e656b1bdb67265fed3d06258ff43ef9: The specified module could not be found.
Which PyMC version is that??
If you're working with Anaconda, it's likely that the PyMC version is outdated. Please follow the installation instructions from our wiki: https://github.com/pymc-devs/pymc/wiki/Installation-Guide-(Windows)
[image: image.png] The version is 3.11.5. I have followed the guide on the wiki. But I still got the error.
On Sun, May 29, 2022 at 5:32 PM Michael Osthege @.***> wrote:
Message sent from a system outside of UConn.
Which PyMC version is that??
If you're working with Anaconda, it's likely that the PyMC version is outdated. Please follow the installation instructions from our wiki: https://github.com/pymc-devs/pymc/wiki/Installation-Guide-(Windows)
— Reply to this email directly, view it on GitHub https://github.com/pymc-devs/pymc/issues/5820#issuecomment-1140528312, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKQHMRW36TJE4OSFU2AVXJTVMPO63ANCNFSM5XIH5TAA . You are receiving this because you authored the thread.Message ID: @.***>
-- Sincerely, Jintao Zhang
Jintao Zhang, Ph.D. candidate Department of Civil & Environmental Engineering University of Connecticut, Storrs, CT 06269
He/Him/His Mobile: (858)247-8401
Please try with the pymc==4.0.0b6 version, or install from the main branch - this is the new version and it's almost released anyhow.
There is an increasing number of installation problems with Theano.
There are multiple closed issues with the same error message: https://github.com/pymc-devs/pymc/issues?q=is%3Aissue+DLL+load+failed+is%3Aclosed
Thank you. Michael. When I update to 4.0.0b6, it does not show this error message. But it shows another error message which I think it's because of the issue with the new version. The error message is displayed below: [image: image.png]
On Tue, May 31, 2022 at 4:59 AM Michael Osthege @.***> wrote:
Message sent from a system outside of UConn.
Please try with the pymc==4.0.0b6 version, or install from the main branch - this is the new version and it's almost released anyhow. There is an increasing number of installation problems with Theano.
There are multiple closed issues with the same error message: https://github.com/pymc-devs/pymc/issues?q=is%3Aissue+DLL+load+failed+is%3Aclosed
— Reply to this email directly, view it on GitHub https://github.com/pymc-devs/pymc/issues/5820#issuecomment-1141864983, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKQHMRSXZTLEXTH6XAPAUTLVMXIF3ANCNFSM5XIH5TAA . You are receiving this because you authored the thread.Message ID: @.***>
-- Sincerely, Jintao Zhang
Jintao Zhang, Ph.D. candidate Department of Civil & Environmental Engineering University of Connecticut, Storrs, CT 06269
He/Him/His Mobile: (858)247-8401
@jiz17043 your image attachment to the email is not visible in GitHub
Sorry, Michael, I have reattached the figure for your reference. Thank you again for your explanation.
Okay, I see.
So first of all it's good that you got a v4 installation.
It looks like the Stochastic Volatility depends on the GaussianRandomWalk which in version 4.0.0b6 was not yet refactored.
But it was migrated just recently. So tit should work with the main branch.
I'm currently re-running it with the latest main and it looks like it's sampling.
If it succeeds I will open a PR over at https://github.com/pymc-devs/pymc-examples/ and ping you there.