I have considered the canceled orders (in the LimitOrders table) simply when cancel = TRUE so they do not include the expired orders.
The issue of the expiration date took to me a few hours as I found expiration dates on the same day of the creation of the order, 5 and 7 days later and other very strange dates such as 31-dec-1969 or 29-October-2035 and I assumed that this diversity, as you say, was due to the use of different applications or different configurations of the user's profile.
My own orders have an expiration date of 31-Dec-1969 (I think it's a typical Unix date) that I think has the "do not expire" effect.
As I was not sure of interpreting this information correctly, I chose not to include it, especially because I thought it was more important to focus on the analysis of the filled orders.
I have considered the canceled orders (in the LimitOrders table) simply when cancel = TRUE so they do not include the expired orders.
The issue of the expiration date took to me a few hours as I found expiration dates on the same day of the creation of the order, 5 and 7 days later and other very strange dates such as 31-dec-1969 or 29-October-2035 and I assumed that this diversity, as you say, was due to the use of different applications or different configurations of the user's profile.
As I was not sure of interpreting this information correctly, I chose not to include it, especially because I thought it was more important to focus on the analysis of the filled orders.
Thanks a lot! @crokkon