Pandas error in documentation notebook
markvdw opened this issue · 6 comments
markvdw commented
Describe the bug
When running the notebook expected_improvement.pct, cell 2 results in an error from pandas noting ValueError: Index data must be 1-dimensional.
The issue seems to be in making a dataframe with a 2D index.
To reproduce
Steps to reproduce the behaviour:
- Download Trieste
- Install Trieste
- Run
expected_improvement.pct
Expected behaviour
The tutorial notebook runs out of the box.
System information
- OS: MacOS
- Python version: 3.10
- Trieste version: Cloned from repo
- TensorFlow version: 2.10.1
- GPflow version: 2.7
The error Traceback is below.
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
Cell In[2], line 9
6 search_space = ScaledBranin.search_space # predefined search space
7 search_space = Box([0, 0], [1, 1]) # define the search space directly
----> 9 fig = plot_function_plotly(
10 scaled_branin,
11 search_space.lower,
12 search_space.upper,
13 grid_density=20,
14 )
15 fig.update_layout(height=400, width=400)
16 fig.show()
File ~/miniconda3/lib/python3.10/site-packages/trieste/experimental/plotting/plotting_plotly.py:252, in plot_function_plotly(obj_func, mins, maxs, grid_density, title, xlabel, ylabel, alpha)
250 for k in range(n_output):
251 f = F[:, k]
--> 252 fig = add_surface_plotly(xx, yy, f, fig, alpha=alpha, figrow=1, figcol=k + 1)
253 fig.update_xaxes(title_text=xlabel, row=1, col=k + 1)
254 fig.update_yaxes(title_text=ylabel, row=1, col=k + 1)
File ~/miniconda3/lib/python3.10/site-packages/trieste/experimental/plotting/plotting_plotly.py:91, in add_surface_plotly(xx, yy, f, fig, alpha, figrow, figcol)
69 def add_surface_plotly(
70 xx: TensorType,
71 yy: TensorType,
(...)
76 figcol: int = 1,
77 ) -> go.Figure:
78 """
79 Adds a surface to an existing plotly subfigure
80
(...)
88 :return: updated plotly figure
89 """
---> 91 d = pd.DataFrame(f.reshape([xx.shape[0], yy.shape[1]]), index=xx, columns=yy)
93 fig.add_trace(
94 go.Surface(z=d, x=xx, y=yy, showscale=False, opacity=alpha, colorscale="viridis"),
95 row=figrow,
96 col=figcol,
97 )
98 return fig
File ~/miniconda3/lib/python3.10/site-packages/pandas/core/frame.py:721, in DataFrame.__init__(self, data, index, columns, dtype, copy)
711 mgr = dict_to_mgr(
712 # error: Item "ndarray" of "Union[ndarray, Series, Index]" has no
713 # attribute "name"
(...)
718 typ=manager,
719 )
720 else:
--> 721 mgr = ndarray_to_mgr(
722 data,
723 index,
724 columns,
725 dtype=dtype,
726 copy=copy,
727 typ=manager,
728 )
730 # For data is list-like, or Iterable (will consume into list)
731 elif is_list_like(data):
File ~/miniconda3/lib/python3.10/site-packages/pandas/core/internals/construction.py:345, in ndarray_to_mgr(values, index, columns, dtype, copy, typ)
335 values = sanitize_array(
336 values,
337 None,
(...)
341 allow_2d=True,
342 )
344 # _prep_ndarraylike ensures that values.ndim == 2 at this point
--> 345 index, columns = _get_axes(
346 values.shape[0], values.shape[1], index=index, columns=columns
347 )
349 _check_values_indices_shape_match(values, index, columns)
351 if typ == "array":
File ~/miniconda3/lib/python3.10/site-packages/pandas/core/internals/construction.py:748, in _get_axes(N, K, index, columns)
746 index = default_index(N)
747 else:
--> 748 index = ensure_index(index)
750 if columns is None:
751 columns = default_index(K)
File ~/miniconda3/lib/python3.10/site-packages/pandas/core/indexes/base.py:7376, in ensure_index(index_like, copy)
7374 return Index._with_infer(index_like, copy=copy, tupleize_cols=False)
7375 else:
-> 7376 return Index._with_infer(index_like, copy=copy)
File ~/miniconda3/lib/python3.10/site-packages/pandas/core/indexes/base.py:717, in Index._with_infer(cls, *args, **kwargs)
715 with warnings.catch_warnings():
716 warnings.filterwarnings("ignore", ".*the Index constructor", FutureWarning)
--> 717 result = cls(*args, **kwargs)
719 if result.dtype == _dtype_obj and not result._is_multi:
720 # error: Argument 1 to "maybe_convert_objects" has incompatible type
721 # "Union[ExtensionArray, ndarray[Any, Any]]"; expected
722 # "ndarray[Any, Any]"
723 values = lib.maybe_convert_objects(result._values) # type: ignore[arg-type]
File ~/miniconda3/lib/python3.10/site-packages/pandas/core/indexes/base.py:541, in Index.__new__(cls, data, dtype, copy, name, tupleize_cols, **kwargs)
538 return cls(arr, dtype, copy=copy, name=name, **kwargs)
540 klass = cls._dtype_to_subclass(arr.dtype)
--> 541 arr = klass._ensure_array(arr, dtype, copy)
542 disallow_kwargs(kwargs)
543 return klass._simple_new(arr, name)
File ~/miniconda3/lib/python3.10/site-packages/pandas/core/indexes/numeric.py:176, in NumericIndex._ensure_array(cls, data, dtype, copy)
172 subarr = data
174 if subarr.ndim > 1:
175 # GH#13601, GH#20285, GH#27125
--> 176 raise ValueError("Index data must be 1-dimensional")
178 subarr = np.asarray(subarr)
179 return subarr
ValueError: Index data must be 1-dimensional
cemkaya-mpi commented
I have the same error. However, it worked with pandas<1.0
ayat-khairy commented
I have the same issue. Any idea how to fix it without downgrading pandas?
uri-granta commented
I'm afraid the notebooks currently don't support pandas 1.0 (or Python 3.10 for that matter). We hope this will be fixed, but it looks as it will be surprisingly fiddly it is not currently viewed as a high priority as it doesn't impact the package itself. Sorry!
hstojic commented
@uri-granta how come notebooks cannot run with 3.10?
uri-granta commented
There are currently three issues to resolve:
- we currently pin sphinx to 3.5.4 to work around plotly/plotly.js#4563, but that version breaks with Python 3.10
- we currently pin pandas to <1.0.0 (and rely on it in a fair few places), which similarly breaks with Python 3.10
- we depend on box2d, which doesn't provide wheels for Python 3.10 or a source package: see https://stackoverflow.com/questions/70323852/error-trying-to-install-python-package-box2d
uri-granta commented