Expected Table-like input error when creating an altair chart

[samesand]

Altair 5.5.0
VegaFusion 2.0.1

I'm trying to create charts using the vegafusion transformer, but I get the following error:
TypeError: Expected Table-like input or dict of arrays or sequence of arrays.

This code generates the error:
import pandas as pd
import altair as alt
chart_data = pd.DataFrame([1,2,3,4,5],columns=['y'])
alt.Chart(chart_data).mark_bar().encode(y='y')

I have a different computer with the same versions of Python, Altair and VegaFusion where this code works.

So far, I have tried to uninstall and reinstall Altair, VegaFusion and Python without success.

[MarcoGorelli]

Hi can you show the full error message please

[samesand]

---

TypeError Traceback (most recent call last)
File c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\site-packages\vegafusion\runtime.py:326, in VegaFusionRuntime._import_inline_datasets(self, inline_datasets, inline_dataset_usage)
324 df_nw = df_nw.select(columns)
--> 326 imported_inline_datasets[name] = Table(df_nw) # type: ignore[arg-type]
327 except TypeError:
328 # Not supported by Narwhals, try pycapsule interface directly

TypeError: Expected Table-like input or dict of arrays or sequence of arrays.

During handling of the above exception, another exception occurred:

TypeError Traceback (most recent call last)
File ~\AppData\Roaming\Python\Python313\site-packages\IPython\core\formatters.py:1036, in MimeBundleFormatter.call(self, obj, include, exclude)
1033 method = get_real_method(obj, self.print_method)
1035 if method is not None:
-> 1036 return method(include=include, exclude=exclude)
1037 return None
1038 else:

File c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\site-packages\altair\vegalite\v5\api.py:3682, in TopLevelMixin.repr_mimebundle(self, *args, **kwds)
3680 else:
3681 if renderer := renderers.get():
-> 3682 return renderer(dct)

File c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\site-packages\altair\utils\display.py:232, in HTMLRenderer.call(self, spec, **metadata)
230 kwargs = self.kwargs.copy()
231 kwargs.update(**metadata, output_div=self.output_div)
--> 232 return spec_to_mimebundle(spec, format="html", **kwargs)

File c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\site-packages\altair\utils\mimebundle.py:129, in spec_to_mimebundle(spec, format, mode, vega_version, vegaembed_version, vegalite_version, embed_options, engine, **kwargs)
127 internal_mode: Literal["vega-lite", "vega"] = mode
128 if using_vegafusion():
--> 129 spec = compile_with_vegafusion(spec)
130 internal_mode = "vega"
132 # Default to the embed options set by alt.renderers.set_embed_options

File c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\site-packages\altair\utils_vegafusion_data.py:273, in compile_with_vegafusion(vegalite_spec)
271 # Pre-evaluate transforms in vega spec with vegafusion
272 row_limit = data_transformers.options.get("max_rows", None)
--> 273 transformed_vega_spec, warnings = vf.runtime.pre_transform_spec(
274 vega_spec,
275 vf.get_local_tz(),
276 inline_datasets=inline_tables,
277 row_limit=row_limit,
278 )
280 # Check from row limit warning and convert to MaxRowsError
281 handle_row_limit_exceeded(row_limit, warnings)

File c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\site-packages\vegafusion\runtime.py:408, in VegaFusionRuntime.pre_transform_spec(self, spec, local_tz, default_input_tz, row_limit, preserve_interactivity, inline_datasets, keep_signals, keep_datasets)
347 """
348 Evaluate supported transforms in an input Vega specification
349
(...)
405 were eligible for pre-transforming
406 """
407 local_tz = local_tz or get_local_tz()
--> 408 imported_inline_dataset = self._import_inline_datasets(
409 inline_datasets, get_inline_column_usage(spec)
410 )
412 new_spec, warnings = self.runtime.pre_transform_spec(
413 spec,
414 local_tz=local_tz,
(...)
420 keep_datasets=parse_variables(keep_datasets),
421 )
423 return new_spec, warnings

File c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\site-packages\vegafusion\runtime.py:330, in VegaFusionRuntime._import_inline_datasets(self, inline_datasets, inline_dataset_usage)
327 except TypeError:
328 # Not supported by Narwhals, try pycapsule interface directly
329 if hasattr(value, "arrow_c_stream"):
--> 330 imported_inline_datasets[name] = Table(value) # type: ignore[arg-type]
331 else:
332 raise

TypeError: Expected Table-like input or dict of arrays or sequence of arrays.

[MarcoGorelli]

ah thanks - this'll be fixed by #560

[yuliap924]

I still have this issue. It happened after I ran alt.data_transformers.enable("vegafusion"):

import numpy as np
import pandas as pd
import altair as alt

alt.data_transformers.enable("vegafusion")
rand = np.random.RandomState(0)
data = pd.DataFrame({'value': rand.randn(100).cumsum()},
index=pd.date_range('2018', freq='D', periods=100))
alt.Chart(data.reset_index()).mark_line().encode(
x='index:T',
y='value:Q'
)

[jonmmease]

Hi @yuliap924, are you on version 2.0.2?

[samesand]

Hello,

I updated to 2.0.2 today and I still get TypeError: Expected Table-like input or dict of arrays or sequence of arrays.

[samesand]

@MarcoGorelli Hello, could you tell me if this is a different issue than the one in #560 ?

[MarcoGorelli]

hi @samesand , I'd have expected it to be the same

the plot you posted renders fine for me

which versions of libraries do you have? could you please show:

import sys
print(sys.executable)
import pandas
import altair
import narwhals
import vegafusion
print(pandas.__file__)
print(pandas.__version__)
print(altair.__file__)
print(altair.__version__)
print(narwhals.__file__)
print(narwhals.__version__)
print(vegafusion.__file__)
print(vegafusion.__version__)

? thanks 🙏

[samesand]

Here is the output:

c:\Users\e193645\AppData\Local\Programs\Python\Python313\python.exe
c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\sitepackages\pandas_init_.py
2.2.3
c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\sitepackages\altair_init_.py
5.5.0
c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\sitepackages\narwhals_init_.py
1.23.0
c:\Users\e193645\AppData\Local\Programs\Python\Python313\Lib\site-packages\vegafusion_init_.py
2.0.2

[MarcoGorelli]

thanks

not sure to be honest - if you make a new virtual environment and install you dependencies afresh in there, does it work?

[samesand]

I tried using a venv and I still get the same error:

TypeError Traceback (most recent call last)
File c:\Users\e193645\OneDrive - \Testing2.venv\Lib\site-packages\vegafusion\runtime.py:326, in VegaFusionRuntime._import_inline_datasets(self, inline_datasets, inline_dataset_usage)
324 df_nw = df_nw.select(columns)
--> 326 imported_inline_datasets[name] = Table(df_nw) # type: ignore[arg-type]
327 except TypeError:
328 # Not supported by Narwhals, try pycapsule interface directly

TypeError: Expected Table-like input or dict of arrays or sequence of arrays.

During handling of the above exception, another exception occurred:

TypeError Traceback (most recent call last)
File c:\Users\e193645\OneDrive - \Testing2.venv\Lib\site-packages\IPython\core\formatters.py:1036, in MimeBundleFormatter.call(self, obj, include, exclude)
1033 method = get_real_method(obj, self.print_method)
1035 if method is not None:
-> 1036 return method(include=include, exclude=exclude)
1037 return None
1038 else:

File c:\Users\e193645\OneDrive - \Testing2.venv\Lib\site-packages\altair\vegalite\v5\api.py:3682, in TopLevelMixin.repr_mimebundle(self, *args, **kwds)
3680 else:
3681 if renderer := renderers.get():
-> 3682 return renderer(dct)

File c:\Users\e193645\OneDrive - \Testing2.venv\Lib\site-packages\altair\utils\display.py:232, in HTMLRenderer.call(self, spec, **metadata)
230 kwargs = self.kwargs.copy()
231 kwargs.update(**metadata, output_div=self.output_div)
--> 232 return spec_to_mimebundle(spec, format="html", **kwargs)

File c:\Users\e193645\OneDrive - \Testing2.venv\Lib\site-packages\altair\utils\mimebundle.py:129, in spec_to_mimebundle(spec, format, mode, vega_version, vegaembed_version, vegalite_version, embed_options, engine, **kwargs)
127 internal_mode: Literal["vega-lite", "vega"] = mode
128 if using_vegafusion():
--> 129 spec = compile_with_vegafusion(spec)
130 internal_mode = "vega"
132 # Default to the embed options set by alt.renderers.set_embed_options

File c:\Users\e193645\OneDrive - \Testing2.venv\Lib\site-packages\altair\utils_vegafusion_data.py:273, in compile_with_vegafusion(vegalite_spec)
271 # Pre-evaluate transforms in vega spec with vegafusion
272 row_limit = data_transformers.options.get("max_rows", None)
--> 273 transformed_vega_spec, warnings = vf.runtime.pre_transform_spec(
274 vega_spec,
275 vf.get_local_tz(),
276 inline_datasets=inline_tables,
277 row_limit=row_limit,
278 )
280 # Check from row limit warning and convert to MaxRowsError
281 handle_row_limit_exceeded(row_limit, warnings)

File c:\Users\e193645\OneDrive - \Testing2.venv\Lib\site-packages\vegafusion\runtime.py:408, in VegaFusionRuntime.pre_transform_spec(self, spec, local_tz, default_input_tz, row_limit, preserve_interactivity, inline_datasets, keep_signals, keep_datasets)
347 """
348 Evaluate supported transforms in an input Vega specification
349
(...) 405 were eligible for pre-transforming
406 """
407 local_tz = local_tz or get_local_tz()
--> 408 imported_inline_dataset = self._import_inline_datasets(
409 inline_datasets, get_inline_column_usage(spec)
410 )
412 new_spec, warnings = self.runtime.pre_transform_spec(
413 spec,
414 local_tz=local_tz,
(...) 420 keep_datasets=parse_variables(keep_datasets),
421 )
423 return new_spec, warnings

File c:\Users\e193645\OneDrive - \Testing2.venv\Lib\site-packages\vegafusion\runtime.py:330, in VegaFusionRuntime._import_inline_datasets(self, inline_datasets, inline_dataset_usage)
327 except TypeError:
328 # Not supported by Narwhals, try pycapsule interface directly
329 if hasattr(value, "arrow_c_stream"):
--> 330 imported_inline_datasets[name] = Table(value) # type: ignore[arg-type]
331 else:
332 raise

TypeError: Expected Table-like input or dict of arrays or sequence of arrays.

[MarcoGorelli]

can you format your code in triple backtricks please, else it's hard/impossible to parse?

[samesand]

my apologies, that was the error output.

This is the code:

import altair as alt
alt.data_transformers.enable('vegafusion')
chart_data = pd.DataFrame([1,2,3,4,5],columns=['y'])
alt.Chart(chart_data).mark_bar().encode(y='y')

[MarcoGorelli]

sure, thanks - and the output from #559 (comment) with your virtual environment activate?

also, could you show what happens if you run

from arro3.core import Table
import pandas as pd
import pyarrow as pa

chart_data = pd.DataFrame([1,2,3,4,5],columns=['y'])
Table(chart_data)

please?