How is Avg Entry Price calculated when using from_orders function?

[jejebelfort]

I am trying to backtest a long-short portfolio where I specify the target weights for different dates.

I am using the following command:
pf = vbt.Portfolio.from_orders(close=prices[size_pct.columns], size=tgt_wgts, group_by=True, lock_cash=True, allow_partial=False, size_type='targetpercent', direction='both', fees=0.0, cash_sharing=True, init_cash=10000)

When checking trades for a specific symbol XYZ, I realize that the Avg Entry price is varying for the same Entry Timestamp and Column (symbol):

      Exit Trade Id Column      Size     Entry Timestamp  Avg Entry Price  Entry Fees      Exit Timestamp  Avg Exit Price  Exit Fees         PnL    Return Direction  Status  Position Id
1968           1968    XYZ  1.083087 2021-01-31 01:00:00       104.056000         0.0 2021-02-05 01:00:00         116.700        0.0  -13.694556 -0.121511     Short  Closed          759
1969           1969    XYZ  0.023349 2021-01-31 01:00:00       108.879749         0.0 2021-02-15 01:00:00         161.014        0.0   -1.217262 -0.478824     Short  Closed          759
1970           1970    XYZ  2.535721 2021-01-31 01:00:00       108.879749         0.0 2021-02-20 01:00:00         235.041        0.0 -319.909686 -1.158721     Short  Closed          759

I would be expecting a same Entry Price for this symbol XYZ. How can I force it to be equal to the input close price on that date 2021-01-31 01:00:00 that I provided in the first input close= ?

PS: I realized there is also a parameter price=, but even setting price=prices[size_pct.columns] it doesn't give me the weights I am looking for and I still observe a different Avg entry price.

PPS: Could you also tell me whether my other parameters are correct (especially lock_cash=True and allow_partial=False given that I want a simple portfolio long short, like the below:

syms=['s1', 's2', 's3', 's4']
weights = [-0.25, -0.25, 0.2, 0.3]

so that abs(weights)=1 in order to have the full amount of cash (init_cash=10000) allocated to those syms given the input prices (no slippage, no fees), say: prices=[1,2,3,4].

In that case I am expecting the allocation to be (in number of asset, not USD):
alloc=np.array([-0.25,-0.25,0.2,0.3])*10000/np.array([1,2,3,4])=[-2500. -1250. 666.66666667 750.]