如何对分类数据和数据框进行预处理

Question

I am preprocessing data for my multiple linear regression model by having a list of genres which I onehotencode

genres = [
    "Action",
    "Adventure",
    "Biography",
    "Comedy",
    "Crime",
    "Erotica",
    "Fantasy",
    "Historical fiction",
    "Horror",
    "Mystery",
    "Romance",
    "Satire",
    "Scifi",
    "Speculative",
    "Thriller",
    "Western",
]

And I also have a user input x_user

x_user = ["Action", "Thriller",]

I want to use x_user as my X_new in:

clf = linear_model.LinearRegression()
clf.fit(X, Y)
clf.predict([X_new])

As I understand it I have to use numeric values when using prediction, so I need to convert x_user to an array X_new with bool as [1,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0]

Is this possible to do with pandas?

I tried

df = pd.DataFrame(data=genres, columns=["genres"])
df['X_new'] = df.genres.apply(lambda q: q.intercept(x_user)).astype(bool)

But got an error

What's the correct way of doing this?

EDIT

My training set looks something like this (after onehotencode)

Y	Action	Adventure	...	Thriller	Western
1.2	1	0	...	1	1
4.7	0	1	...	1	0
...	...	...	...	...	...

And the test set is from a user input and looks something like (after onehotencode)

Action	Thriller
1	1

But I want it to look like this

Action	Adventure	...	Thriller	Western
1	0	...	1	0

Answer 1

Could it be something as simple as:

X_new = pd.DataFrame(0, index=np.arange(len(x_user)), columns=genres)
X_new['Action'] = x_user.Action
X_new['Thriller'] = x_user.Thriller

assuming x_user is a pandas dataframe. If it is instead a numpy array or something similar just assign it based on its index.

If you manage the columns of x_user so that they are in the correct order you could generalise to:

X_new[['Action','Thriller']] = x_user

If x_user is a pandas dataframe:

X_new[X_user.columns] = x_user.values