sklearn linear regression is doesn't appear to fit correctly

Question

I'm putting together a simple OLS example from sklearn, and I'm noticing strange results. Below is the results

from sklearn.pipeline import Pipeline
from sklearn.linear_model import LinearRegression

model = Pipeline([('linear', LinearRegression(fit_intercept=True))])

n = 100
x = np.linspace(0, 10, n)
eps = np.random.randn(n)
y = 0.5 * x + -2.5 + eps

model.fit(y.reshape(-1, 1), x.reshape(-1, 1))
yhat = model.predict(x.reshape(-1, 1))
plt.scatter(x, y)
plt.plot(x, yhat, 'r')

It seems strange that the OLS fit is so off. Just looking for someone else to reproduce this before posting this on the main sklearn issue tracker. My versions are below

sklearn=0.22.1
python=3.6.1

Answer 1

Here is where you did wrong

model.fit(y.reshape(-1, 1), x.reshape(-1, 1))

change to:

model.fit(x.reshape(-1, 1), y.reshape(-1, 1))