python-scikit-learnHow to inverse transform PCA reduced data
from sklearn import decomposition, datasets
X, y = datasets.load_iris(return_X_y=True)
pca = decomposition.PCA(n_components=3)
pca.fit(X)
reduced = pca.transform(X)
inversed = pca.inverse_transform(reduced)ctrl + cfrom sklearn importimport module from scikit-learn | load_irisloads Iris dataset |
decomposition.PCA(create PCA dimensionality reduction model | n_componentsreduce to the given number of components (3 in our case) |
.fit(train reduction model model | .transform(transform original data and return reduced dimensions data |
.inverse_transform(inverse transform back to original number of dimensions | |
Usage example
from sklearn import decomposition, datasets
X, y = datasets.load_iris(return_X_y=True)
pca = decomposition.PCA(n_components=3)
pca.fit(X)
reduced = pca.transform(X)
inversed = pca.inverse_transform(reduced)
print('Reduced: ', reduced.shape)
print('Invresed: ', inversed.shape)
print(inversed[0:5])output
Reduced: (150, 3)
Invresed: (150, 4)
[[5.09928623 3.50072335 1.40108561 0.1982949 ]
[4.86875839 3.03166108 1.4475168 0.12536791]
[4.69370023 3.20638436 1.30958161 0.18495067]
[4.6238432 3.07583667 1.46373578 0.25695828]
[5.0193263 3.58041421 1.37060574 0.24616799]]
More of Python Scikit Learn
- Grid search CV example
- Using OneClassSVM for outliers detection example
- SVC classifier example
- Using Ridge regression model example
- How to get best estimator from grid search CV
- Use imputer for one column only
- How to get best parameters from grid search CV
- PCA dimensionality reduction example
- Decision tree classifier example
See more codes...