forest-confidence-interval icon indicating copy to clipboard operation
forest-confidence-interval copied to clipboard

Published 20 hours ago • verified publisher icon scikit-learn-contrib

Overflow errors

Open tawe141 opened this issue 5 years ago • 5 comments

When using random_forest_error() with a dataset in which the features range between 0 and 1 and of datatype float64, I get a bunch of overflow errors like so:

/Users/erictaw/forest-confidence-interval/forestci/calibration.py:86: RuntimeWarning: overflow encountered in exp
  g_eta_raw = np.exp(np.dot(XX, eta)) * mask
/Users/erictaw/forest-confidence-interval/forestci/calibration.py:101: RuntimeWarning: overflow encountered in exp
  g_eta_raw = np.exp(np.dot(XX, eta_hat)) * mask
/Users/erictaw/forest-confidence-interval/forestci/calibration.py:102: RuntimeWarning: invalid value encountered in true_divide
  g_eta_main = g_eta_raw / sum(g_eta_raw)

Turning off calibration eliminates these errors, of course. Is this something I should be worried about?

tawe141 avatar Mar 03 '20 00:03 tawe141

I've been getting this error and have the same question please?

PrenilS avatar Mar 30 '20 11:03 PrenilS

When using random_forest_error() with a dataset in which the features range between 0 and 1 and of datatype float64, I get a bunch of overflow errors like so:

/Users/erictaw/forest-confidence-interval/forestci/calibration.py:86: RuntimeWarning: overflow encountered in exp
  g_eta_raw = np.exp(np.dot(XX, eta)) * mask
/Users/erictaw/forest-confidence-interval/forestci/calibration.py:101: RuntimeWarning: overflow encountered in exp
  g_eta_raw = np.exp(np.dot(XX, eta_hat)) * mask
/Users/erictaw/forest-confidence-interval/forestci/calibration.py:102: RuntimeWarning: invalid value encountered in true_divide
  g_eta_main = g_eta_raw / sum(g_eta_raw)

Turning off calibration eliminates these errors, of course. Is this something I should be worried about?

I have the same probrem anthe the errors are gone after turning off calibration. Have you found other solutions?

haijunli0629 avatar Dec 24 '20 07:12 haijunli0629

When using random_forest_error() with a dataset in which the features range between 0 and 1 and of datatype float64, I get a bunch of overflow errors like so:

/Users/erictaw/forest-confidence-interval/forestci/calibration.py:86: RuntimeWarning: overflow encountered in exp
  g_eta_raw = np.exp(np.dot(XX, eta)) * mask
/Users/erictaw/forest-confidence-interval/forestci/calibration.py:101: RuntimeWarning: overflow encountered in exp
  g_eta_raw = np.exp(np.dot(XX, eta_hat)) * mask
/Users/erictaw/forest-confidence-interval/forestci/calibration.py:102: RuntimeWarning: invalid value encountered in true_divide
  g_eta_main = g_eta_raw / sum(g_eta_raw)

Turning off calibration eliminates these errors, of course. Is this something I should be worried about?

@tawe141

When turning off calibration, the V_IJ_unbias array will contain negetive values, which was mentioned in #25 . If not, all the output is NaN. Do you have any solutions to this?

Thanks.

haijunli0629 avatar Dec 25 '20 14:12 haijunli0629

I am still experiencing this issue.

sylphrena0 avatar Jul 08 '22 16:07 sylphrena0