SKLearn Linear Module
- The SK learn linear module is one such module that helps to study the relationship between the independent and dependent variables.
- The linear module can be implemented by using the best line.
- The module used for using linear regression is “sklearn.linear_module.LinearRegression”.
- The linear regression module is used to minimize the sum of the squares of residual between the objects in a set by fitting a linear model.
Parameters
- fit_intercept: Boolean, default = True
- Normalize: Boolean, default = False
- Copy_x: Boolean, default = True
- N_jobs: int default = None
- There are four parameters used by the linear regression module, they are:
- fit_intercept: Boolean, default = True
- If the fit_intercept parameter is set with false, then there will be no intercept.
- The fit_intercept parameter is used to find the intercept of a model.
- Normalize: Boolean, default = False
- If the fit_intercept is false, then the normalized parameter will be ignored.
- If the normalized parameter is defined with true, then normalization of regressor x(independent variable) is done before the normalization of regression.
- The x regressor is normalized by dividing the mean with L2 form after subtracting the mean.
- Copy_x: Boolean, default = True
- If the copy_x parameter is set to false, then the x will be overwritten.
- But for the copy_x parameter, its default value is set to true, which means the value x can be copied.
- N_jobs: int default = None
- Here the n_job parameter is used for computation.
- It is used to speed up in large problems, which is if first n_objects is greater than 1 and x is parse which is set to true.
Attributes:
- rank_: int: It is only used when the x is denser. Which is used to find the rank of the x matrix?
- Singular_: array of shape (min (x, y)): It is only used when the x is denser. Which is used to find the singular values of the x matrix?
- Coef_: array of shape (n_features,): In regression problems are the estimated coefficients.
If there are many targets in fitting,it is a 2d array when only 1 target is given, which is a 1 d array of length. - Intercept_:array of shape(n_targets,): If the fit_intercept is equal to false, then it is set to 0.0, It is an independent object in the model, which is linear.
- n_features_in_: int: It is no of the features found in fit.
- Feature_names_in_:ndarray of shape(n_features_in_,): It is no of the features found in fit.If x has only all strings.
Example:
>>> import numpy as np
>>> from sklearn.linear_model import Linearregression
>>> x = np.array([[1, 1], [1, 2], [2, 2], [2, 3]] )
>>> y = np.dot (x, np.array ([ 1, 2] )) +3
>>>regr = linearregression( ).fit (x, y)
>>>Regr.score (x, y)
1.0
>>>Regr.coef_
array ([1., 2.])
>>>regr.intercept_
3.0 . . .
>>>regr.predict (np.array([[3, 5]]))
array ([16. ])
Methods:
fit(x, y, sample_weight = none)
- The parameters used are x, y, and sample_weight.
- Here x means training data.
- Y means target values. It is sometimes cast to x if needed.
- Sample_weight means the individual weight of every sample. Its default value is None.
- This method returns an estimator that is fitted.
Get_params (deep = true)
- The parameters used are deep.
- The deep means if it is set to true, then it returns the estimator and subobjects that are contained.
- The Deep default value is set to true.
- This method returns parameter which are mapped to values.
Predict(x)
- The parameters used are x;
- Here the x means samples.
- This method returns values that are predicted.
Scores(x, y, sample_weight = none)
- The parameters used in this method are x, y, andsample_weight.
- x means test samples.
- Here y means values of x which are true.
- Here sample_weight means sample weights.
- This method returns the score:float.
R2 of self. Predict with respect to y.
- Here the r2 is used when the score o the regressor is called. That is when the regressor uses multioutput = “uniform_average”.
Set_params( **params)
- This method is used to work on simple estimators along with objects which are nested.
- It can be possible to update the component of the nested object because its parameters in the form of < component >__< parameter >
- The parameter used is **params.
- Params mean estimator parameters.
- This method returns the estimator instance.