eMouse drift output cluster metrics

[GrimmSnark]

Hello,

I am playing with the eMouse simulation to test out some parameters. The simulation has a benchmark script and output (output_clusters_metrics.txt) to compare to ground truth. See below for sample of this output. However I can not find any explanation of the various measures calcuated.

Does anyone happen to have more insight into this?

GTlabel	nSpike	meanAmp	meanPos	undetected	bestMiss	bestFP	bestScore	autoMiss	autoFP	autoScore	nMerges	phy labels
1	10273	79.587	214.411	0.061	0.0716	0.002	0.9264	0.0716	0.002	0.9264	0	42
2	1326	110.717	606.248	0.006	0.0098	0.2331	0.7571	0.1048	0.2516	0.6436	1	131 130
3	9680	60.734	418.576	0.869	0.9997	0.9816	-0.9813	0.9997	0.9816	-0.9813	0	103
4	11349	125.849	134.111	0.363	0.4184	0.0003	0.5813	0.4184	0.0003	0.5813	0	33
5	2382	133.14	250	0.018	0.0218	0.803	0.1752	0.0218	0.803	0.1752	0	53

All the best,

Michael