predict_search

astropy.vo.client.conesearch.predict_search(url, *args, **kwargs)

Predict the run time needed and the number of objects for a Cone Search for the given access URL, position, and radius.

Run time prediction uses astropy.utils.timer.RunTimePredictor. Baseline searches are done with starting and ending radii at 0.05 and 0.5 of the given radius, respectively.

Extrapolation on good data uses least-square straight line fitting, assuming linear increase of search time and number of objects with radius, which might not be accurate for some cases. If there are less than 3 data points in the fit, it fails.

Warnings (controlled by warnings) are given when:

1. Fitted slope is negative.
2. Any of the estimated results is negative.
3. Estimated run time exceeds astropy.utils.data.Conf.remote_timeout.

Note

If verbose=True, extra log info will be provided. But unlike conesearch_timer(), timer info is suppressed.

If plot=True, plot will be displayed. Plotting uses matplotlib.

The predicted results are just rough estimates.

Prediction is done using conesearch(). Prediction for AsyncConeSearch is not supported.

Parameters:

url : str

Cone Search access URL to use.

args, kwargs : see conesearch()

Extra keyword plot is allowed and only used by this function and not conesearch().

Returns:

t_est : float

Estimated time in seconds needed for the search.

n_est : int

Estimated number of objects the search will yield.

Raises:

AssertionError

If prediction fails.

ConeSearchError

If input parameters are invalid.

VOSError

If VO service request fails.