Tell me please:
There are a number y ~ i (1), i.e. integrated first order (non-stationary). To bring to stationarity I use the increment of logarithm. The problem is that when the density function is displayed, according to Oy, where there should be probabilities in decimal form more than one. With a nuclear assessment, the same. The teacher wants the increment of logarithm.
import numpy as np Import schipy.stats AS STS np.random.seed (123) y = np.linspace (50,200,150) + np.random.normal (Loc = 0, Scale = 15, Size = 150) x = np.diff (np.log (y)) MU, SIGMA = STS.NORM.FIT (X) xmin, xmax = x.min (), x.max () x_l = np.linspace (Xmin, Xmax, Len (X)) DENS = STS.NORM.PDF (X_L, LOC = MU, SCALE = SIGMA) Plt.plot (X_L, DENS) DENS_KDE = STS.GAUSSIAN_KDE (X) DENS_PDF = DENS_KDE.EVALUATE (X_L) Plt.plot (X_L, DENS_PDF)
Answer 1, Authority 100%
sts.norm.pdf probability density function (probability density function) can return values more than one, but if it is integrated, the value is close to one:
in : from scipy.integrarate import simps In : SIMPS (DENS, X = X_L) Out : 0.999763132933244