ValueError: The cpd states do not match expected states: expected {0, 1}, found {'yes', 'no'}

[rafmora]

Description

Thank YOU guys for the good work in developing this excellent API. Very useful
I ran the tutorial https://causalnex.readthedocs.io/en/latest/03_tutorial/03_tutorial.html, but I am getting an error. I do not know what I am not doing correctly.

Context

Tutorial

Steps to Reproduce

steps from the tutorial

Expected Result

the cps is expecting states 0,1; but it sees Yes, No

Actual Result

Stops when running
print("distribution before do", ie.query()["higher"])
ie.do_intervention("higher",
{'yes': 1.0,
'no': 0.0})
print("distribution after do", ie.query()["higher"])

-- If you received an error, place it here.

ValueError Traceback (most recent call last)
in
2 ie.do_intervention("higher",
3 {'yes': 1.0,
----> 4 'no': 0.0})
5 print("distribution after do", ie.query()["higher"])

C:\ProgramData\Anaconda3\lib\site-packages\causalnex\inference\inference.py in do_intervention(self, node, state)
211 state = {s: float(s == state) for s in self._cpds[node]}
212
--> 213 self._do(node, state)
214 self._generate_bbn()
215

C:\ProgramData\Anaconda3\lib\site-packages\causalnex\inference\inference.py in _do(self, observation, state)
173 "The cpd states do not match expected states: expected {expected}, found {found}".format(
174 expected=set(self._cpds_original[observation]),
--> 175 found=set(state.keys()),
176 )
177 )

ValueError: The cpd states do not match expected states: expected {0, 1}, found {'yes', 'no'}
-- Separate them if you have more than one.

## Your Environment


* CausalNex version used 0.9.1
* Python version used (`python -V`): Python 3.7.9
* Operating system and version: Windows 10

[PNEPNE]

Hi @rafmora,

I tried to run this recently and the output looks alright on my machine (see screenshot below). If you are running from a Jupyter notebook, perhaps you can try to restart the kernel and re-run? It might also be worth reinstalling/upgrading python and CausalNex just in case.

PS: For the benefit of anyone else reading through this, please refer to our tutorial for full codes - https://causalnex.readthedocs.io/en/stable/03_tutorial/03_tutorial.html

[oentaryorj]

Closing this issue as the do-intervention seems to be working fine, as per @PNEPNE's example. Feel free to raise a new issue if the problem persists. Thank you.