Book a Demo!
CoCalc Logo Icon
StoreFeaturesDocsShareSupportNewsAboutPoliciesSign UpSign In
holoviz
GitHub Repository: holoviz/panel
 Path: blob/main/examples/apps/django_multi_apps/gbm/pn_model.py
2014 views
1
import hvplot

2
import hvplot.pandas

3
import numpy as np

4
import pandas as pd

5
import param

6


7
import panel as pn

8


9


10
class GBM(param.Parameterized):

11
    # interface

12
    mean = param.Number(default=5, bounds=(.0, 25.0))

13
    volatility = param.Number(default=5, bounds=(.0, 50.0))

14
    maturity = param.Integer(default=1, bounds=(0, 25))

15
    n_observations = param.Integer(default=10, bounds=(2, 100))

16
    n_simulations = param.Integer(default=20, bounds=(1, 500))

17
    refresh = pn.widgets.Button(name="Refresh", button_type='primary')

18


19
    def __init__(self, **params):

20
        super().__init__(**params)

21


22
    # update the plot for every changes

23
    # @param.depends('mean', 'volatility', 'maturity', 'n_observations', 'n_simulations', watch=True)

24
    # refresh the plot only on button refresh click

25
    @param.depends('refresh.clicks')

26
    def update_plot(self, **kwargs):

27
        df_s = pd.DataFrame(index=range(0, self.n_observations))

28


29
        for s in range(0, self.n_simulations):

30
            name_s = f"stock_{s}"

31
            df_s[name_s] = self.gbm(spot=100,

32
                                    mean=self.mean/100,

33
                                    vol=self.volatility/100,

34
                                    dt=self.maturity / self.n_observations,

35
                                    n_obs=self.n_observations)

36


37
        return df_s.hvplot(grid=True, colormap='Paired', value_label="Level", legend=False)

38


39
    @staticmethod

40
    def gbm(spot: float, mean: float, vol: float, dt: float, n_obs: int) -> np.ndarray:

41
        """ Geometric Brownian Motion

42


43
        :param spot: spot value

44
        :param mean: mean annualised value

45
        :param vol: volatility annualised value

46
        :param dt: time steps

47
        :param n_obs: number of observation to return

48
        :return: a geometric brownian motion np.array()

49
        """

50
        # generate normal random

51
        rand = np.random.standard_normal(n_obs)

52
        # initialize the parameters

53
        S = np.zeros_like(rand)

54
        S[0] = spot

55
        # loop to generate the brownian motion

56
        for t in range(1, n_obs):

57
            S[t] = S[t - 1] * np.exp((mean - (vol ** 2) / 2) * dt + vol * rand[t] * np.sqrt(dt))

58
        # return the geometric brownian motion

59
        return S

60


61