Automatic calibration not giving a correct output

[DaniAlcAlv]

The water calibration output JSON does not compute the regression slope and offset (intercept) correctly. The plot below shows a comparison between the recomputed fit (red line), using the correctly calculated values, and the original fit from the output (orange dotted line):

Also, the output interval_average dictionary values are not dividing the water_weight values by the repeat_count as they should.

[DaniAlcAlv]

Ideally, the automatic calibration should call the exact same code used for the manual calibration, so both methods remain consistent and neither becomes outdated if there are changes:
https://github.com/AllenNeuralDynamics/Aind.Behavior.Services/blob/main/src/aind_behavior_services/rig/water_valve.py