Discretization of Continuous Variables in Pymc

Questions v5
7

The CPT to be defined for the “combination” node:

file_path = 'Combination.xlsx'
df = pd.read_excel(file_path, sheet_name = 0, header=None)  
Combination_cpt = df.to_numpy().T
Combination_cpt.shape

image
image973×231 18.5 KB

node definition：

    PoA = pm.Triangular('PoA', lower=0, c=0.3, upper=1)
    CF = pm.Lognormal('CF', mu=4.5, sigma=0.2)
    TEF = pm.Deterministic('TEF', CF * PoA)
    MPLEF = pm.Deterministic('MPLEF', TEF * vulnerability)
    CoSL = pm.Normal('CoSL', mu=0.7, sigma=0.2)
    PLF = pm.Poisson('PLF', mu=MPLEF)
    SLF = pm.Binomial('SLF', n=PLF, p=CoSL)
# The "combination" node is a child node of both PLF and SLF, and the CPT data is imported from Excel. 
# It is desired to divide both PLF and SLF into 12 intervals each to match the dimensions of CPT