SR.jl to include the constant length, variables to be present in equations and previous state information to be used in current state

[pravin-pandey]

Hello, everyone. I am using SR.jl for fluid dynamics application and need my SR to work with some constraints:

1. The constants generated are currently in the range of e-16, and the value is 0. I want to constrain SRRegressor so that the constant value should be of a particular length. Also, the equations primarily use constants rather than variables.
2. I would like all the variables in equations since all the inputs impact the output feature. Currently, only 1 or 2 variables are used in the equations. A lower loss value is achieved but hardly has any input variable, for example:

Complexity,Loss,Equation
18,1.5375358344591114e-7,"(0.003040617289178417 + 0.003040617289178417) / (sqrt(x₁) + ((2.009394600234112 + x₁) + 0.003040617289178417))"

3. Is there any feature in SR.jl with previous state information to predict the current state, somewhat like an RNN?

I am new to SR and Julia, so if there are any redundant questions, please let me know the most relevant solution.