Hashing¶

class
category_encoders.hashing.
HashingEncoder
(max_process=0, max_sample=0, verbose=0, n_components=8, cols=None, drop_invariant=False, return_df=True, hash_method='md5')[source]¶ A multivariate hashing implementation with configurable dimensionality/precision.
The advantage of this encoder is that it does not maintain a dictionary of observed categories. Consequently, the encoder does not grow in size and accepts new values during data scoring by design.
It’s important to read about how max_process & max_sample work before setting them manually, inappropriate setting slows down encoding.
Default value of ‘max_process’ is 1 on Windows because multiprocessing might cause issues, see in : https://github.com/scikitlearncontrib/categoricalencoding/issues/215 https://docs.python.org/2/library/multiprocessing.html?highlight=process#windows
 Parameters
 verbose: int
integer indicating verbosity of the output. 0 for none.
 cols: list
a list of columns to encode, if None, all string columns will be encoded.
 drop_invariant: bool
boolean for whether or not to drop columns with 0 variance.
 return_df: bool
boolean for whether to return a pandas DataFrame from transform (otherwise it will be a numpy array).
 hash_method: str
which hashing method to use. Any method from hashlib works.
 max_process: int
how many processes to use in transform(). Limited in range(1, 64). By default, it uses half of the logical CPUs. For example, 4C4T makes max_process=2, 4C8T makes max_process=4. Set it larger if you have a strong CPU. It is not recommended to set it larger than is the count of the logical CPUs as it will actually slow down the encoding.
 max_sample: int
how many samples to encode by each process at a time. This setting is useful on low memory machines. By default, max_sample=(all samples num)/(max_process). For example, 4C8T CPU with 100,000 samples makes max_sample=25,000, 6C12T CPU with 100,000 samples makes max_sample=16,666. It is not recommended to set it larger than the default value.
 n_components: int
how many bits to use to represent the feature. By default we use 8 bits. For highcardinality features, consider using upto 32 bits.
References
 1
Feature Hashing for Large Scale Multitask Learning, from
https://alex.smola.org/papers/2009/Weinbergeretal09.pdf .. [R8dde675226a22] Don’t be tricked by the Hashing Trick, from https://booking.ai/dontbetrickedbythehashingtrick192a6aae3087
Methods
fit
(X[, y])Fit encoder according to X and y.
fit_transform
(X[, y])Fit to data, then transform it.
Returns the names of all transformed / added columns.
get_params
([deep])Get parameters for this estimator.
hashing_trick
(X_in[, hashing_method, N, …])A basic hashing implementation with configurable dimensionality/precision
set_params
(**params)Set the parameters of this estimator.
transform
(X[, override_return_df])Call _transform() if you want to use single CPU with all samples
require_data

fit
(X, y=None, **kwargs)[source]¶ Fit encoder according to X and y.
 Parameters
 Xarraylike, shape = [n_samples, n_features]
Training vectors, where n_samples is the number of samples and n_features is the number of features.
 yarraylike, shape = [n_samples]
Target values.
 Returns
 selfencoder
Returns self.

get_feature_names
()[source]¶ Returns the names of all transformed / added columns.
 Returns
 feature_names: list
A list with all feature names transformed or added. Note: potentially dropped features are not included!

static
hashing_trick
(X_in, hashing_method='md5', N=2, cols=None, make_copy=False)[source]¶ A basic hashing implementation with configurable dimensionality/precision
Performs the hashing trick on a pandas dataframe, X, using the hashing method from hashlib identified by hashing_method. The number of output dimensions (N), and columns to hash (cols) are also configurable.
 Parameters
 X_in: pandas dataframe
description text
 hashing_method: string, optional
description text
 N: int, optional
description text
 cols: list, optional
description text
 make_copy: bool, optional
description text
 Returns
 outdataframe
A hashing encoded dataframe.
References
Cite the relevant literature, e.g. [R6b702480991a1]. You may also cite these references in the notes section above. .. [R6b702480991a1] Kilian Weinberger; Anirban Dasgupta; John Langford; Alex Smola; Josh Attenberg (2009). Feature Hashing for Large Scale Multitask Learning. Proc. ICML.