Using the KDE class¶

Notebook Setup¶

In [1]:

import os

if ("COLAB_RELEASE_TAG" in os.environ):
    !pip install ssm-simulators

import os if ("COLAB_RELEASE_TAG" in os.environ): !pip install ssm-simulators

In [2]:

if any(["MARIMO" in item_ \
    for item_ in os.environ]):
    import marimo as mo
    mo.dependencies(pip=["ssm-simulators",
                         "pandas",
                         "matplotlib"
                         ])

if any(["MARIMO" in item_ \ for item_ in os.environ]): import marimo as mo mo.dependencies(pip=["ssm-simulators", "pandas", "matplotlib" ])

Bare-bones Tutorial¶

In [3]:

import ssms
from ssms.support_utils.kde_class import LogKDE
import numpy as np
from matplotlib import pyplot as plt

import ssms from ssms.support_utils.kde_class import LogKDE import numpy as np from matplotlib import pyplot as plt

In [4]:

sim_out = ssms.basic_simulators.simulator.simulator(model = 'ddm',
                                                    theta = dict(v = 1.0, a = 1.5, z = 0.5, t = 0.4),
                                                    n_samples = 50000
                                                    )

sim_out = ssms.basic_simulators.simulator.simulator(model = 'ddm', theta = dict(v = 1.0, a = 1.5, z = 0.5, t = 0.4), n_samples = 50000 )

In [5]:

my_kde = LogKDE(sim_out,
                bandwidth_type = "silverman",
                auto_bandwidth = True,
                displace_t = False)

my_kde = LogKDE(sim_out, bandwidth_type = "silverman", auto_bandwidth = True, displace_t = False)

In [6]:

rt_data = {'rts': np.repeat(np.linspace(0, 10, 1000), 2),
           'choices': np.tile([-1, 1], 1000)}

rt_data = {'rts': np.repeat(np.linspace(0, 10, 1000), 2), 'choices': np.tile([-1, 1], 1000)}

In [7]:

test_lb = -66.6
kde_out = my_kde.kde_eval(data = rt_data,
                          log_eval = False)

test_lb = -66.6 kde_out = my_kde.kde_eval(data = rt_data, log_eval = False)

In [8]:

plt.plot(rt_data['rts'][rt_data['choices'] == 1], 
         kde_out[rt_data['choices'] == 1])
plt.plot(rt_data['rts'][rt_data['choices'] == -1], 
         kde_out[rt_data['choices'] == -1])
plt.xlim(0,10)

plt.plot(rt_data['rts'][rt_data['choices'] == 1], kde_out[rt_data['choices'] == 1]) plt.plot(rt_data['rts'][rt_data['choices'] == -1], kde_out[rt_data['choices'] == -1]) plt.xlim(0,10)

Out[8]:

(0.0, 10.0)