Error TypeError: __init__() got an unexpected keyword argument 'X'

[YoizNunez]

Hello Dear,

Could you help me review this error, in its latest version of the library it does not allow me to use it correctly.

I use this example:

estimators = pymint.load_models()
X,y = pymint.load_data()
X = X.astype({'urban': 'category', 'rural':'category'})

explainer = pymint.InterpretToolkit(estimators,X=X, y=y,)

[monte-flora]

Are you using the latest version (0.2.5)? In an earlier version of the code, X and y were referred to by different names. But I changed over to X and y syntax to be consistent with scikit-learn. I apologize for not including a deprecation notice. I was able to pip install the latest version and the above code segment works on my end. If you continue to see the error, please let me know.

[YoizNunez]

Thank you for your reply! I appreciate! I will try to run the library later today. Best regards, Yoiz Nuñez I was using the version 0.2.4.

…

On Mon, Feb 7, 2022, 8:53 AM Montgomery Flora ***@***.***> wrote: Are you using the latest version (0.2.5)? In an earlier version of the code, X and y were referred to by different names. But I changed over to X and y syntax to be consistent with scikit-learn. I apologize for not including a deprecation notice. I was able to pip install the latest version and the above code segment works on my end. If you continue to see the error, please let me know. — Reply to this email directly, view it on GitHub <#73 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ARHP3XFGX34HTPCWKRXZLYDUZ7MGDANCNFSM5NVQJNUQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[monte-flora]

Were you successful? I recently deleted older versions of the package, which we unstable, and used the old syntax. I also made some updates and released version 0.2.6.

[YoizNunez]

Hi Dear, Thank you for your email. Currently I am getting this error when I try to install the package py-mint (I am using the google colab environement): !pip install py-mint Error shown: Collecting py-mint Downloading py_mint-0.2.5-py2.py3-none-any.whl (35.1 MB) |████████████████████████████████| 35.1 MB 1.8 MB/s Requirement already satisfied: scikit-learn in /usr/local/lib/python3.7/dist-packages (from py-mint) (1.0.2) Requirement already satisfied: tqdm in /usr/local/lib/python3.7/dist-packages (from py-mint) (4.62.3) Requirement already satisfied: pandas in /usr/local/lib/python3.7/dist-packages (from py-mint) (1.3.5) Requirement already satisfied: statsmodels in /usr/local/lib/python3.7/dist-packages (from py-mint) (0.10.2) Collecting shap>=0.30.0 Downloading shap-0.40.0-cp37-cp37m-manylinux2010_x86_64.whl (564 kB) |████████████████████████████████| 564 kB 28.7 MB/s Requirement already satisfied: xarray>=0.16.0 in /usr/local/lib/python3.7/dist-packages (from py-mint) (0.18.2) Requirement already satisfied: matplotlib in /usr/local/lib/python3.7/dist-packages (from py-mint) (3.2.2) Collecting py-mint Downloading py_mint-0.2.4-py2.py3-none-any.whl (35.1 MB) |████████████████████████████████| 35.1 MB 1.3 MB/s Downloading py_mint-0.2.3-py2.py3-none-any.whl (35.1 MB) |████████████████████████████████| 35.1 MB 121 kB/s Downloading py_mint-0.2.2-py2.py3-none-any.whl (35.1 MB) |████████████████████████████████| 35.1 MB 294 kB/s Downloading py_mint-0.2.1-py2.py3-none-any.whl (35.1 MB) |████████████████████████████████| 35.1 MB 174 kB/s Downloading py_mint-0.2.0-py2.py3-none-any.whl (35.1 MB) |████████████████████████████████| 35.1 MB 220 kB/s ERROR: Cannot install py-mint==0.2.0, py-mint==0.2.1, py-mint==0.2.2, py-mint==0.2.3, py-mint==0.2.4 and py-mint==0.2.5 because these package versions have conflicting dependencies. The conflict is caused by: py-mint 0.2.5 depends on stanford-theme py-mint 0.2.4 depends on stanford-theme py-mint 0.2.3 depends on stanford-theme py-mint 0.2.2 depends on stanford-theme py-mint 0.2.1 depends on stanford-theme py-mint 0.2.0 depends on stanford-theme To fix this you could try to: 1. loosen the range of package versions you've specified 2. remove package versions to allow pip attempt to solve the dependency conflict ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/user_guide/#fixing-conflicting-dependencies El mié, 16 feb 2022 a las 11:02, Montgomery Flora ***@***.***>) escribió:

…

Were you successful? I recently deleted older versions of the package, which we unstable, and used the old syntax. I also made some updates and released version 0.2.6. — Reply to this email directly, view it on GitHub <#73 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ARHP3XENLZS75LELDNQWEJ3U3PKARANCNFSM5NVQJNUQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

-- Yoiz Nuñez Research Assistant Microwave Laboratory CETUC / PUC Rio

[monte-flora]

What version of pip are you using? There was an issue with pip installing previous versions of a package (pypa/pip#9284). I would explicitly pip install py-mint==0.2.6 to avoid that issue.

[YoizNunez]

Ok. I used pip install py-mint==0.2.6 in the anaconda environment, and the package was successfully installed. However, only for testing, in the google colaboratory, for one reason the command !pip install py-mint==0.2.6 returns an error of package installation. ERROR: Could not find a version that satisfies the requirement py-mint==0.2.6 (from versions: 0.2.4, 0.2.5) ERROR: No matching distribution found for py-mint==0.2.6 El jue, 17 feb 2022 a las 12:51, Montgomery Flora ***@***.***>) escribió:

…

What version of pip are you using? There was an issue with pip installing previous versions of a package (pypa/pip#9284 <pypa/pip#9284>). I would explicitly pip install py-mint==0.2.6 to avoid that issue. — Reply to this email directly, view it on GitHub <#73 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ARHP3XEALQDRKFWESP4Q5QLU3U7RPANCNFSM5NVQJNUQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

-- Yoiz Nuñez Research Assistant Microwave Laboratory CETUC / PUC Rio

[monte-flora]

I do not have much experience with google colaboratory. Does it require updating? (since it is seeing the older versions?)

[YoizNunez]

It could be that. I will review this issue, because sometimes I use that environment too. :) Thank you for your help. Best regards, El jue, 17 feb 2022 a las 13:22, Montgomery Flora ***@***.***>) escribió:

…

I do not have much experience with google colaboratory. Does it require updating? (since it is seeing the older versions?) — Reply to this email directly, view it on GitHub <#73 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ARHP3XHHM22ETIXWWRAAHSDU3VDFPANCNFSM5NVQJNUQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

-- Yoiz Nuñez Research Assistant Microwave Laboratory CETUC / PUC Rio

[YoizNunez]

Hi Dear, I have a question. In your trained model, that you use as examples. How to normalize categorical entries. Do you use some kind of OneHotEncoder in your code? It's only for doing the same procedure in my models before to use the library py-mint. El jue, 17 feb 2022 a las 15:09, Yoiz Nuñez ***@***.***>) escribió:

…

It could be that. I will review this issue, because sometimes I use that environment too. :) Thank you for your help. Best regards, El jue, 17 feb 2022 a las 13:22, Montgomery Flora (< ***@***.***>) escribió: > I do not have much experience with google colaboratory. Does it require > updating? (since it is seeing the older versions?) > > — > Reply to this email directly, view it on GitHub > <#73 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ARHP3XHHM22ETIXWWRAAHSDU3VDFPANCNFSM5NVQJNUQ> > . > Triage notifications on the go with GitHub Mobile for iOS > <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> > or Android > <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. > > You are receiving this because you authored the thread.Message ID: > ***@***.***> > -- Yoiz Nuñez Research Assistant Microwave Laboratory CETUC / PUC Rio

-- Yoiz Nuñez Research Assistant Microwave Laboratory CETUC / PUC Rio

[monte-flora]

It is not common to normalize categorical features since categorical features are usually ordinal or nominal data types. If you are using mixed data types (numerical and categorical) and want to have the numeric features normalized, I recommend using the column transformer method in scikit-learn (https://scikit-learn.org/stable/auto_examples/compose/plot_column_transformer_mixed_types.html)

For the dataset used in the tutorial notebooks, the two categorical features (rural and urban) are binary (0,1) variables so no encoding was necessary. All that is necessary is to specify they are categorical features in the pandas.DataFrame.

[YoizNunez]

Thank you very much for your reply. Perfect. My question was if you were using encoding for the categorical variables. I understand that for your dataset it was not necessary. So, I am using the same Pipeline too. [image: Pipeline.png] El mar, 8 mar 2022 a las 15:52, Montgomery Flora ***@***.***>) escribió:

…

It is not common to normalize categorical features since categorical features are usually ordinal or nominal data types. If you are using mixed data types (numerical and categorical) and want to have the numeric features normalized, I recommend using the column transformer method in scikit-learn ( https://scikit-learn.org/stable/auto_examples/compose/plot_column_transformer_mixed_types.html ) For the dataset used in the tutorial notebooks, the two categorical features (rural and urban) are binary (0,1) variables so no encoding was necessary. All that is necessary is to specify they are categorical features in the pandas.DataFrame. — Reply to this email directly, view it on GitHub <#73 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ARHP3XALP4DRTCZLE3H7A23U67D23ANCNFSM5NVQJNUQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

-- Yoiz Nuñez Research Assistant Microwave Laboratory CETUC / PUC Rio

[monte-flora]

Glad I could clarify! Feel free to ask additional questions, if necessary.

[YoizNunez]

Thank you! I appreciate your help! Best Regards, Yoiz Nuñez El mar, 8 mar 2022 a las 16:08, Montgomery Flora ***@***.***>) escribió:

…

Glad I could clarify! Feel free to ask additional questions, if necessary. — Reply to this email directly, view it on GitHub <#73 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ARHP3XE5HFSMAAQFU3MLWWDU67FW3ANCNFSM5NVQJNUQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

-- Yoiz Nuñez Research Assistant Microwave Laboratory CETUC / PUC Rio

[YoizNunez]

FYI...

---------- Forwarded message --------- De: Yoiz Nuñez ***@***.***> Date: mar, 8 mar 2022 a las 16:13 Subject: Re: [monte-flora/py-mint] Error TypeError: __init__() got an unexpected keyword argument 'X' (Issue #73) To: monte-flora/py-mint < ***@***.***> Cc: monte-flora/py-mint ***@***.***>, Author < ***@***.***> Thank you! I appreciate your help! Best Regards, Yoiz Nuñez El mar, 8 mar 2022 a las 16:08, Montgomery Flora ***@***.***>) escribió:

Glad I could clarify! Feel free to ask additional questions, if necessary. — Reply to this email directly, view it on GitHub <#73 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ARHP3XE5HFSMAAQFU3MLWWDU67FW3ANCNFSM5NVQJNUQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

Hi Dear, One more doubt. I am new to training and interpreting machine learning models. I performed the normalization on my dataset, considering for example a numeric input ("frequency") and in my case I am using OneHotEncoding for a categorical input called number of walls that has 3 classes (1 wall, 2 walls, 3 walls). When performing the encoding of that categorical variable, it is transformed into 3 columns with binary combinations. So my normalized and transformed dataset would be formed by Frequency Wall 1 Wall 2 Walls 3 0.5 1 0 0 0.2 0 1 0 0.1 0 0 1 ..... That same dataset that I used to train the model (4 inputs) I must use in the Py-mint library, right, indicating only that Wall1, Wall2 and Wall 3 are categorical variables. Is this correct? Thank you for your help!

…

-- Yoiz Nuñez Research Assistant Microwave Laboratory CETUC / PUC Rio

[YoizNunez]

Hi Dear, One more doubt. I am new to training and interpreting machine learning models. I performed the normalization on my dataset, considering for example a numeric input ("frequency") and in my case I am using OneHotEncoding for a categorical input called number of walls that has 3 classes (1 wall, 2 walls, 3 walls). When performing the encoding of that categorical variable, it is transformed into 3 columns with binary combinations. So my normalized and transformed dataset would be formed by Frequency Wall 1 Wall 2 Walls 3 0.5 1 0 0 0.2 0 1 0 0.1 0 0 1 ..... That same dataset that I used to train the model (4 inputs) I must use in the Py-mint library, right, indicating only that Wall1, Wall2 and Wall 3 are categorical variables. Is this correct? Thank you for your help! El mar, 8 mar 2022 a las 16:08, Montgomery Flora ***@***.***>) escribió:

…

Glad I could clarify! Feel free to ask additional questions, if necessary. — Reply to this email directly, view it on GitHub <#73 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ARHP3XE5HFSMAAQFU3MLWWDU67FW3ANCNFSM5NVQJNUQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

-- Yoiz Nuñez Research Assistant Microwave Laboratory CETUC / PUC Rio