Hi,
I was using the rename_checkpoint script and noticed that there was an issue that occurred when appending data that changes order of magnitude.
For example if the simulation was stopped at 95s and the restarted simulation ran to a time of 150s it would only append data where the recorded time is less than 100s.
The function appends all data from the checkpoint stat files where the recorded time is 'larger' than the last time on the original stat file. The issue is that the time is stored in standard form and concat_statfile extracts and stores time as a string, it then evaluates using this string. When logically comparing strings python would determine the following to be true:
"0.99979164887571699E+002" > "0.10004628322103638E+003"
Hi,
I was using the rename_checkpoint script and noticed that there was an issue that occurred when appending data that changes order of magnitude.
For example if the simulation was stopped at 95s and the restarted simulation ran to a time of 150s it would only append data where the recorded time is less than 100s.
The function appends all data from the checkpoint stat files where the recorded time is 'larger' than the last time on the original stat file. The issue is that the time is stored in standard form and concat_statfile extracts and stores time as a string, it then evaluates using this string. When logically comparing strings python would determine the following to be true:
"0.99979164887571699E+002" > "0.10004628322103638E+003"