#!/usr/bin/env python
# coding: utf-8

# ### non ipympl backends
# 
# every function in this library should work with any interactive matplotlib backend. Although the functions from `mpl_interactions.jupyter` assume that you are in a notebook context in order to display the sliders.

# In[ ]:


# NBVAL_SKIP
get_ipython().run_line_magic('matplotlib', 'qt5')


# In[ ]:


import numpy as np

import matplotlib.pyplot as plt
from mpl_interactions import interactive_plot


# The below cell will display the sliders in the notebook but a separate matplotlib window will pop up. An important caveat is that the performance of `interactive_plot` seems to be significantly worse with a non-ipympl backend.

# In[ ]:


x = np.linspace(0, np.pi, 100)
τ = np.linspace(1, 10, 100)
β = np.linspace(1, 10, 100)


def f(x, τ, β):
    return np.sin(x * τ) * x ** β


fig, ax = plt.subplots()
controls = interactive_plot(x, f, τ=τ, β=β)


# ### generic submodule
# 
# `mpl_interactions.generic` contains functions that will work in any matplotlib context. So for example the below code will work in the notebook or from a script

# In[ ]:


from mpl_interactions.generic import heatmap_slicer


# In[ ]:


x = np.linspace(0, np.pi, 100)
y = np.linspace(0, 10, 200)
X, Y = np.meshgrid(x, y)
data1 = np.sin(X) + np.exp(np.cos(Y))
data2 = np.cos(X) + np.exp(np.sin(Y))
fig, axes = heatmap_slicer(
    x,
    y,
    (data1, data2),
    slices="both",
    heatmap_names=("dataset 1", "dataset 2"),
    labels=("Some wild X variable", "Y axis"),
    interaction_type="move",
)


# In[ ]: