Can u add an example where we can use the data in numpy for the generated "tpot_mnist_pipeline.py"

[anjith2006]

In the example MNIST, the generated python code is not directly usable for evaluation with the original
data.
Can you post an example were I can directly use the scikit type , or simply numpy arrays for data and labels.

[rhiever]

Do you mean this example? Here's an example of that code transforming the data from load_digits() into pandas DataFrame format.

import numpy as np
import pandas as pd

from sklearn.cross_validation import StratifiedShuffleSplit
from sklearn.linear_model import LogisticRegression
from sklearn.datasets import load_digits

digits = load_digits()

tpot_data = pd.DataFrame(digits.data)
tpot_data['class'] = digits.target

training_indeces, testing_indeces = next(iter(StratifiedShuffleSplit(tpot_data['class'].values, n_iter=1, train_size=0.75, test_size=0.25)))

result1 = tpot_data.copy()

# Perform classification with a logistic regression classifier
lrc1 = LogisticRegression(C=2.8214285714285716)
lrc1.fit(result1.loc[training_indeces].drop('class', axis=1).values, result1.loc[training_indeces, 'class'].values)
result1['lrc1-classification'] = lrc1.predict(result1.drop('class', axis=1).values)

[pronojitsaha]

Hey @rhiever , we need to change the StratifiedShuffleSplit to train_test_split in the above tpot example as well.

[rhiever]

Sure, submit the PR? :-)

On Thursday, March 17, 2016, PRONOjit Saha notifications@github.com wrote:

Hey @rhiever https://github.com/rhiever , we need to change the
StratifiedShuffleSplit to train_test_split in the above tpot example as
well.

—
You are receiving this because you were mentioned.
Reply to this email directly or view it on GitHub
#114 (comment)

Randal S. Olson, Ph.D.
Postdoctoral Researcher, Institute for Biomedical Informatics
University of Pennsylvania

E-mail: rso@randalolson.com | Twitter: @randal_olson
https://twitter.com/randal_olson
http://www.randalolson.com

[pronojitsaha]

OK...will look into it and send across the PR.