metrics.significance2

Methods to calculate the significance

significance2 Module

class metrics.significance2.Significance(tmp_dir, output_dir)

Class for significance calculation

checkSignificance(input, result_file, q1=0.050000000000000003, q2=0.94999999999999996, check_value=0)

Checks if a value is statistically significant different from zero Uses the bootstrapped files

Parameters:

• input – list with files or folder of bootstraps
• result_file – file to check

Returns:

sig_lon,sig_lat lists with significant points

save_significance_mask(sign_lon, sign_lat, result_file)

Saves the significance mask in a netcdf file

Parameters:

• sign_lon,sign_lat – significant points
• result_file – File to overwrite

_Significance__getLonLat(ifile)

Get vectors with lon and lat values from a netdf file using cdo.griddes Was introduced because we were using a damaged grid lon,lat valued are located in the center of a gridbox

Parameters:ifile – netcdf fn Result :lon,lat vectors

_Significance__getQuantile(mVar, q1, q2, precision=1.0)

Returns the q’th percentile of the distribution given in the argument ‘data’. Uses the ‘precision’ parameter to control the noise level.

Parameters:

• mvar – numpy variable
• q1,q2 – lower and upper level

Returns:

q’th percentile