Updating graph during algorithms

[alkngo]

After each group of orders is made (the for loop), live._live_send_back (shown below) should send the updated list of buys and sells to each clients to update the graphs. In effect, the clients' graphs would be updating periodically during the algorithm.

Instead, it seems that all live._live_send_back calls don't run until the entire algorithm ends (the while loop), which results in the clients' graph not updating until the entire algorithm is finished.

https://github.com/Leeps-Lab/flow_market/blob/17c1407214ab67b5aa6dde119efc92bec69c1e7b/models.py#L97-L140

I use a similar method to input orders through a file (shown below), but that method runs as expected, so I'm not sure what's causing the issue for the algorithm methods.

https://github.com/Leeps-Lab/flow_market/blob/17c1407214ab67b5aa6dde119efc92bec69c1e7b/models.py#L175-L198

[giangd]

I'm not sure what the correct behavior should look like, so I'm going to tag this for further discussion with Yilin and Alex.

[giangd]

How algorithmic inputs should work

• 50 volume at 6 units at a time
• Immediately put into market 6 units for buy/sell
• When market executes that order, algorithm immediately places another 6 unit for buy/sell while:
  • While time (specified by user) has not expired
  • While less than total volume
• The 6 units have lowest time priority (since they were added recently)

Required parameters

• Total volume (exists already) <--- no, it doesn't!
• Executed units (new)
• Units at a time (new)
• Expiration time (new)
• [ ] Units at a time is indicated by q_max
• [ ] New parameter: "q_total"

Todo

Buys

• [x] Puts order back into market while (total executed units < total units)
• [ ] The order should have lowest time priority
  • [ ] Add flag to add at end
• [x] Deletes order when timer is up

Sells

• [x] Puts order back into market while (total executed units < total units)
• [ ] The order should have lowest time priority
  • [ ] Add flag to add at end
• [x] Deletes order when timer is up