search

SuperDARNCanada / fitacf.3.0

The repo for the new and improved fitacf routine
GNU General Public License v3.0
0 stars 1 forks source link

Added the fixed code that addresses the incorrect treatment of salt/pepper data #11

Closed kkotyk closed 7 years ago

kkotyk commented 7 years ago

Fixed that uninitialized variable issue.

pasha-ponomarenko commented 7 years ago

In the same commit the cut-off threshold for lag 0 power has been changed from three sigmas above the noise mean, cutoff_pwr = noise_pwr (1. + NOISE_SIGMA_THRESHOLD/sqrt(fit_prms->nave)), which allows analysing some of the data with SNR <1, to two time noise mean, cutoff_pwr = noise_pwr 2 which corresponds to SNR=1. This modification was supposed to be done in a separate pull request but has been inintentionally performed in this one due to my mistake. The reason is that the three-sigma threshold allowed an excessive amount of isolated noise "spikes" to be present in the data between the bands of real scatter. We still need to decide if we leave the threshold fixed at SNR=1 or make it flexible. Right now I tend to leave it as it is because it makes the results more compatible with the older FITACF versions, and more stringent filtering can always be performed afterwards if necessary.