search

mapa17 / Eruliaf

A BitTorrent like overlay peer simulator
GNU General Public License v3.0
1 stars 0 forks source link

Xrange in statistics too high #66

Closed mapa17 closed 12 years ago

mapa17 commented 12 years ago

For some strange reasons the generated statistics are generated over the whole simulation scenario now, even though there is no data available to visualize.

mapa17 commented 12 years ago

NAN values are appended to the end of the generated data and even if they are not printed they still scale the x axis.

mapa17 commented 12 years ago

The error is inside the simulation, as peers even they have downloaded all peers, dont detect that they finished the download!

Because of that they stay in the simulation and the strange statistics bug is produced!

mapa17 commented 12 years ago

The source of the problem does not seems to be peers not finishing their downloads. This seems to work.

The NaN values in the pData table ( in the Statistics.R script ) come to be, because there is no incomplete peer after some time. If than the mean value and similar is calculated, it will result in an NaN.

Strange, because this must have always been like that. Why is the xaxis now scaled by that? The bug will be fixed by calculating the limits on my own and than specify them as x-axis limits.