Could you please provide the order in which one should go through the topics?

We haven't run these examples in almost two years. Would be good to check for updates.

您好，关于书中使用的数据集，可否告知出处，在书中没有找到数据集的链接。

The bottom three images' code is inserted in a text box instead of a code box.

Fix:
pip install --upgrade matplotlib

If you cant update MPL , this will fix the issue:

ax.set_yscale('log', nonposy='clip')

Notebooks are missing a requirements.txt for dependency version control

Hi everyone,
there are many datasets used in this book, but I didn't see source page links attached in the content. Could someone share those dataset please? Thank you!

Hi Alice,

Thanks for the great book! There are two lines of code in the last two notebooks for the end-to-end examples that I don't quite understand:

• if order < 5: order += 1
• if order < top_n: order += 1

What was the rationale behind the fact that you used if blocks here?

Thank you!

#Hello,

Jupyter notebooks have a links to data files in local directories, but there are no directories in the repository, so it is unable to run notebooks, what is correct way to run these notebooks, need to manually download data files?

Thank you.

can you provide URL or dataset for feature engineering book

I downloaded the mag_papers_0 file from the following link, https://www.openacademic.ai/oag/ , but didn't find the mag_subset20K.txt used in https://github.com/alicezheng/feature-engineering-book/blob/master/09.01-05_%5BEnd-to-End_Example%5D_Recommender_Take_1.ipynb.

Is the mag_subset20K.txt file reconstructed from one of the source file, if so, how to build? if not, where is the right link to download this file?

Thanks

hi, when i run the 04.01-06 sample code,

Create the target column--True for Nightlife businesses, and False otherwise

twobiz_reviews['target'] = twobiz_reviews.apply(lambda x: 'Nightlife' in x['categories'],
axis=1)
there is a error:
"ValueError: Cannot set a frame with no defined index and a value that cannot be converted to a Series"
and i use the yelp dataset v13
how to solve this?

I refer you to the fit method of the KMeansFeaturizer class for the code listing mentioned in the subject.
When the pretrain approach was used (i.e. including information from the class labels), the second re run of the k-means algorithm to get the clusters in the original space was initialized with the init argument as init=km_model_pretrain.cluster_centers_[:,:2].
I think this will only work with a two feature dataset and will break for a dataset with additional features.
Can you please clarify your approach.
Thanks

Hello, Where can i find hash_fcn() and sign_hash() in 5-3 and 5-4?