# EphemeralDataContext' object has no attribute 'sources'
**URL:** https://discourse.greatexpectations.io/t/ephemeraldatacontext-object-has-no-attribute-sources/2042
**Category:** GX Core Support
**Created:** [December 29, 2024, 9:45pm UTC](https://discourse.greatexpectations.io/t/ephemeraldatacontext-object-has-no-attribute-sources/2042 "2024-12-29T21:45:54Z")
**Posts on this page:** 2
**Page:** 1
### Author:  [@resolver101757](https://discourse.greatexpectations.io/u/resolver101757)
#### Post date: [December 29, 2024, 9:45pm UTC](https://discourse.greatexpectations.io/t/ephemeraldatacontext-object-has-no-attribute-sources/2042/1 "2024-12-29T21:45:54Z")
In the documenation, it says to `If you have a configured Data Source, Data Asset, and Batch Definition you can test your Expectation before adding it to your Expectation Suite. To do this see Test an Expectation.`. However, when i create batch data from my pandas dataframe i get the error EphemeralDataContext. What am i doing wrong?
`--------------------------------------------------------------------------- AttributeError Traceback (most recent call last) Cell In[57], [line 2](vscode-notebook-cell:?execution_count=57&line=2) [1](vscode-notebook-cell:?execution_count=57&line=1) # Create a datasource from the pandas DataFrame ----> [2](vscode-notebook-cell:?execution_count=57&line=2) datasource = context.sources.pandas_default.read_dataframe( [3](vscode-notebook-cell:?execution_count=57&line=3) df_sample, [4](vscode-notebook-cell:?execution_count=57&line=4) name="ms_marco_datasource" [5](vscode-notebook-cell:?execution_count=57&line=5) ) AttributeError: 'EphemeralDataContext' object has no attribute 'sources'`
Hers the code to recreate the problem:
````import
# Create reproducible sample data
sample_data = {
'query': [
'what is the capital of france?',
'how tall is mount everest?',
'who wrote romeo and juliet?',
'what year did world war 2 end?',
'what is the speed of light?'
],
'answers': [
['Paris'],
['29,029 feet (8,848 meters)'],
['William Shakespeare'],
['1945'],
['299,792,458 meters per second']
],
'query_type': [
'LOCATION',
'NUMERIC',
'PERSON',
'NUMERIC',
'NUMERIC'
],
'query_id': [1, 2, 3, 4, 5],
'passages': [
{'is_selected': [1], 'passage_text': ['Paris is the capital of France.'], 'url': ['example.com/1']},
{'is_selected': [1], 'passage_text': ['Mount Everest is 29,029 feet tall.'], 'url': ['example.com/2']},
{'is_selected': [1], 'passage_text': ['Romeo and Juliet was written by William Shakespeare.'], 'url': ['example.com/3']},
{'is_selected': [1], 'passage_text': ['World War 2 ended in 1945.'], 'url': ['example.com/4']},
{'is_selected': [1], 'passage_text': ['The speed of light is 299,792,458 meters per second.'], 'url': ['example.com/5']}
],
'wellFormedAnswers': [
['Paris is the capital of France.'],
['Mount Everest is 29,029 feet (8,848 meters) tall.'],
['William Shakespeare wrote Romeo and Juliet.'],
['World War 2 ended in 1945.'],
['The speed of light is 299,792,458 meters per second.']
]
}
df_sample = pd.DataFrame(sample_data)
print("\nSample dataset:")
df_sample
import great_expectations as gx
context = gx.get_context()
# Create a datasource from the pandas DataFrame
datasource = context.sources.pandas_default.read_dataframe(
df_sample,
name="ms_marco_datasource"
)
data_asset_name = "my_data_asset"
batch_definition_name = "my_batch_definition"
batch_definition = (
context.data_sources.get(df_pd_ms_macro)
.get_asset(data_asset_name)
.get_batch_definition(batch_definition_name)
) ```
````
Without batched data everything runs fine but from what i can see i need batch data to validate, can someone help me out to what the solution is?
---
### Author:  [@joshstauffer](https://discourse.greatexpectations.io/u/joshstauffer)
#### Post date: [March 12, 2025, 2:08pm UTC](https://discourse.greatexpectations.io/t/ephemeraldatacontext-object-has-no-attribute-sources/2042/2 "2025-03-12T14:08:09Z")
Hey @resolver101757, there are a couple issues with your code.
> [@resolver101757](#):
>
> ```auto
> datasource = context.sources.pandas_default.read_dataframe(
> df_sample,
> name="ms_marco_datasource"
> )
>
> ```
As the error says, there isn’t a `sources` attribute on the `DataContext`, it should be `data_sources`.
> [@resolver101757](#):
>
> ```auto
> datasource = context.sources.pandas_default.read_dataframe(
> df_sample,
> name="ms_marco_datasource"
> )
>
> ```
This method returns a `Batch`, not a data source, and accepts an `asset_name` parameter, not a `name` parameter. You can test expectations directly against this `Batch` by calling `batch.validate(expect)`. Here’s an updated version of your code:
```auto
sample_data = {
'query': [
'what is the capital of france?',
'how tall is mount everest?',
'who wrote romeo and juliet?',
'what year did world war 2 end?',
'what is the speed of light?'
],
'answers': [
['Paris'],
['29,029 feet (8,848 meters)'],
['William Shakespeare'],
['1945'],
['299,792,458 meters per second']
],
'query_type': [
'LOCATION',
'NUMERIC',
'PERSON',
'NUMERIC',
'NUMERIC'
],
'query_id': [1, 2, 3, 4, 5],
'passages': [
{'is_selected': [1], 'passage_text': ['Paris is the capital of France.'], 'url': ['example.com/1']},
{'is_selected': [1], 'passage_text': ['Mount Everest is 29,029 feet tall.'], 'url': ['example.com/2']},
{'is_selected': [1], 'passage_text': ['Romeo and Juliet was written by William Shakespeare.'], 'url': ['example.com/3']},
{'is_selected': [1], 'passage_text': ['World War 2 ended in 1945.'], 'url': ['example.com/4']},
{'is_selected': [1], 'passage_text': ['The speed of light is 299,792,458 meters per second.'], 'url': ['example.com/5']}
],
'wellFormedAnswers': [
['Paris is the capital of France.'],
['Mount Everest is 29,029 feet (8,848 meters) tall.'],
['William Shakespeare wrote Romeo and Juliet.'],
['World War 2 ended in 1945.'],
['The speed of light is 299,792,458 meters per second.']
]
}
import pandas as pd
df_sample = pd.DataFrame(sample_data)
import great_expectations as gx
import great_expectations.expectations as gxe
context = gx.get_context()
# Create a batch from the pandas DataFrame
batch = context.data_sources.pandas_default.read_dataframe(df_sample)
expect = gxe.ExpectColumnToExist(column="query_id")
result = batch.validate(expect=expect)
print(result.describe())
```