Estimator (main class)

class scitime.RuntimeEstimator(meta_algo='RF', verbose=3, confidence=0.95)

Bases: scitime._log.LogMixin

Estimator class arguments

Parameters:
  • meta_algo – meta algorithm (RF or NN)
  • verbose – log output (0, 1, 2 or 3)
  • confidence – confidence for intervals
time(algo, X, y=None)

predicts training runtime for a given training

Parameters:
  • algo – algo whose runtime the user wants to predict
  • X – np.array of inputs to be trained
  • y – np.array of outputs to be trained (set to None is unsupervised algo)
Returns:

predicted runtime, low and high values of the confidence interval
Return type:

float