Thank you very much for your patient explanation! Based on my re-reading of your answer and the documentation, I used the following code to plot and obtain the data units after being read by happi:

S = happi.Open(data_path, verbose=False, pint=True)
screen_diag = S.Screen(
    diagNumber=0,
    timesteps=None,
    units={"x": "MeV", "y": "um"}
)
screen_diag.plot()

The effect is shown in the figure.

According to Document ParticleBinning units, my understanding is that, based on the data read here, if I want to convert it to the number of real particles per unit z-direction, I only need to multiply by a conversion factor from weight to the number of real particles per unit z-direction, which is N_r * (L_r)^d, where N_r is the reference density, L_r is the reference length, and d is the dimension of the simulation. Is my understanding correct?

I specifically told you not to do that first. I recommend you use your own conversion for now.

Thank you very much for your patient reply! I have compared the direct unit conversion of the happi package with the manual conversion of read data, and found that their units are consistent. The results of the two methods are shown in the figure below.


Here, the first figure shows the direct conversion using the happi package, while the second figure presents the result of manual unit conversion after reading the data manually.
I am still more concerned about one point: I only need to multiply by the conversion coefficient that converts the weight per unit in the z-direction to the actual number of particles, which is N_r * (L_r)^d. Here, N_r is the reference density, L_r is the reference length, and d is the simulation dimension. I would like to confirm whether my understanding of the calculation of this conversion coefficient is correct.

Thank you so much for your warm and patient explanations—I have fully understood the unit conversion issue for the screen! The unit conversion here is equivalent to taking the values read by happi, multiplying them by Δa × ΔE, and then multiplying the result by Nr × Lr^d (where Nr stands for reference density, Lr for reference length, and d for simulation dimension) expressed in the units I need.