Version 0.00s9 Graph of proportion DFLE/TLE when popbased=1 added

Version 0.00s9 Graph of proportion DFLE/TLE when popbased=1 added

If T_....txt file was correct. The graph wasn't plotted when
popbased=1 (thanks to Saito). Also, the y2tics as well as the label
"popbased=" were not unset and all graphs generated after the graph of
proportions were displayed with these wrong information.

The difference between verion 0.99s7 and 0.99s8 resides when firstpass is higher than 1 and when
there is no interview at that firstpass and later because the death
occurred between the first wave and firstpass: a death event was added
to the likelihood, instead of skipping that person. I only recently
read that the ChangeLog can be verbose...