I am facing this issue while importing lazypredict

[iammuhammadnoumankhan]

Actually, i was importing lazy predict version 0.2.12 and sklearn==1.5.1

i got this issue

---

TypeError Traceback (most recent call last)
Cell In[4], line 1
----> 1 from lazypredict.Supervised import LazyClassifier
2 from sklearn.model_selection import train_test_split

File ~\anaconda3\envs\base_env\Lib\site-packages\lazypredict\Supervised.py:98
89 # CLASSIFIERS.append(('CatBoostClassifier',catboost.CatBoostClassifier))
91 numeric_transformer = Pipeline(
92 steps=[("imputer", SimpleImputer(strategy="mean")), ("scaler", StandardScaler())]
93 )
95 categorical_transformer_low = Pipeline(
96 steps=[
97 ("imputer", SimpleImputer(strategy="constant", fill_value="missing")),
---> 98 ("encoding", OneHotEncoder(handle_unknown="ignore", sparse=False)),
99 ]
100 )
102 categorical_transformer_high = Pipeline(
103 steps=[
104 ("imputer", SimpleImputer(strategy="constant", fill_value="missing")),
(...)
107 ]
108 )
111 # Helper function

TypeError: OneHotEncoder.init() got an unexpected keyword argument 'sparse'

[qzc438]

I have the same issue.

[lesshaste]

It seems to be a scikit learn version problem. See https://stackoverflow.com/a/77997362/1473517 . Lazypredict needs updating but I don't know if anyone is still maintaining.

[iammuhammadnoumankhan]

Actually the problem is with the library, in Pypl, I mean from where we are installing the library it's * .12 and the problem is in that version, when I checked on Github, the latest version as *.10 and then I checked the code it was correct. And the version *.10 is not available to install with pip. I mean in Pypl.

…

On Sun, 22 Sept 2024, 9:40 pm lesshaste, ***@***.***> wrote: It seems to be a scikit learn version problem. See https://stackoverflow.com/a/77997362/1473517 . Lazypredict needs updating but I don't know if anyone is still maintaining. — Reply to this email directly, view it on GitHub <#456 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AZ2V2JWTYYOUG4HBHCDKNHDZX3XHFAVCNFSM6AAAAABOAINYHOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNRWHA3DKMZZGU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[iammuhammadnoumankhan]

For a solution, I downgraded sklearn by a few versions of this to 1.4.* if I'm not wrong. On Mon, 23 Sept 2024, 12:29 am Muhammad Nouman Khan, < ***@***.***> wrote:

…

Actually the problem is with the library, in Pypl, I mean from where we are installing the library it's * .12 and the problem is in that version, when I checked on Github, the latest version as *.10 and then I checked the code it was correct. And the version *.10 is not available to install with pip. I mean in Pypl. On Sun, 22 Sept 2024, 9:40 pm lesshaste, ***@***.***> wrote: > It seems to be a scikit learn version problem. See > https://stackoverflow.com/a/77997362/1473517 . Lazypredict needs > updating but I don't know if anyone is still maintaining. > > — > Reply to this email directly, view it on GitHub > <#456 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AZ2V2JWTYYOUG4HBHCDKNHDZX3XHFAVCNFSM6AAAAABOAINYHOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNRWHA3DKMZZGU> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> >

[MrIsCM]

Hello,

The issue is related to Scikit-learn version changes. From scikit-learn documentation:

Added in version 1.2: sparse was renamed to sparse_output

It appears they changed the keyword 'sparse' to 'sparse_output'.
The solution would simply be to change the old keyword to the new one.

Edit:
It looks like this has been fixed in the code (I see an update 7 months ago).
However, the latest release was 28th September 2022.

[CoryKornowicz]

I got past this issue by installing from git through pip.
pip install git+https://github.com/shankarpandala/lazypredict.git
This grabbed the latest fixes that support the new "sparse_output" naming.

[jorobledo]

Worst case, just open the file in your error message (~\anaconda3\envs\base_env\Lib\site-packages\lazypredict\Supervised.py:98):

File ~\anaconda3\envs\base_env\Lib\site-packages\lazypredict\Supervised.py:98
89 # CLASSIFIERS.append(('CatBoostClassifier',catboost.CatBoostClassifier))
91 numeric_transformer = Pipeline(
92 steps=[("imputer", SimpleImputer(strategy="mean")), ("scaler", StandardScaler())]
93 )
95 categorical_transformer_low = Pipeline(
96 steps=[
97 ("imputer", SimpleImputer(strategy="constant", fill_value="missing")),
---> 98 ("encoding", OneHotEncoder(handle_unknown="ignore", sparse=False)),

And errase the , sparse=False from line 98.

Then just restart your kernel if you are within jupyter notebook, and voilá