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

# In[1]:


get_ipython().run_line_magic('load_ext', 'autoreload')
get_ipython().run_line_magic('autoreload', '2')

get_ipython().run_line_magic('matplotlib', 'inline')

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt


# In[2]:


plt.rcParams['figure.figsize'] = [14, 9]


# In[20]:


from sklearn.linear_model import LogisticRegression
from sklearn.datasets import load_iris


# In[24]:


data = load_iris()
X, y = data.data, data.target
names = data.feature_names
est = LogisticRegression(solver='lbfgs', multi_class='auto')
est.fit(X, y)


# In[30]:


from sklearn.ensemble import RandomForestClassifier


# In[33]:


dtree = RandomForestClassifier(n_estimators=100)
dtree.fit(X, y)


# # First Plot

# In[34]:


from sklearn_plot_api import plot_partial_dependence
features = [0, 1, 2, 3]
viz = plot_partial_dependence(est, X, features, feature_names=[names[i] for i in features])


# In[35]:


from sklearn_plot_api import plot_partial_dependence
features = [0, 1, 2, 3]
viz = plot_partial_dependence(dtree, X, features, feature_names=[names[i] for i in features])


# ## Adjust color of sepal width

# In[37]:


viz.artists_['sepal width (cm)'].set_color('green')
viz.figure_


# ## Add y label to petal length

# In[38]:


viz.axes_['petal length (cm)'].set_ylabel('Parital dependence')
viz.figure_


# ## Passing bounds through axes

# In[41]:


from sklearn_plot_api import plot_partial_dependence
fig, (ax1, ax2) = plt.subplots(1, 2)
features = [0, 1, 2, 3]
viz1 = plot_partial_dependence(est, X, features, feature_names=[names[i] for i in features], ax=ax1)
viz2 = plot_partial_dependence(dtree, X, features, feature_names=[names[i] for i in features], ax=ax2)