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

# # 天气最高温度预测任务

# In[2]:


# Pandas is used for data manipulation
import pandas as pd

# Read in data as pandas dataframe and display first 5 rows
features = pd.read_csv('data/temps.csv')
features.head(5)


# In[9]:


features_list = list(features.columns)
features_list


# In[11]:


features_list.index('temp_1')


# # 数据大小

# In[4]:


print('The shape of our features is:', features.shape)


# In[5]:


# Descriptive statistics for each column
features.describe()


# 无异常数据，可以直接用啦

# In[6]:


# Use datetime for dealing with dates
import datetime

# Get years, months, and days
years = features['year']
months = features['month']
days = features['day']

# List and then convert to datetime object
dates = [str(int(year)) + '-' + str(int(month)) + '-' + str(int(day)) for year, month, day in zip(years, months, days)]
dates = [datetime.datetime.strptime(date, '%Y-%m-%d') for date in dates]


# In[7]:


dates[:5]


# # 数据展示

# In[8]:


# Import matplotlib for plotting and use magic command for Jupyter Notebooks
import matplotlib.pyplot as plt

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

# Set the style
plt.style.use('fivethirtyeight')


# In[9]:


# Set up the plotting layout
fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(nrows=2, ncols=2, figsize = (10,10))
fig.autofmt_xdate(rotation = 45)

# Actual max temperature measurement
ax1.plot(dates, features['actual'])
ax1.set_xlabel(''); ax1.set_ylabel('Temperature'); ax1.set_title('Max Temp')

# Temperature from 1 day ago
ax2.plot(dates, features['temp_1'])
ax2.set_xlabel(''); ax2.set_ylabel('Temperature'); ax2.set_title('Previous Max Temp')

# Temperature from 2 days ago
ax3.plot(dates, features['temp_2'])
ax3.set_xlabel('Date'); ax3.set_ylabel('Temperature'); ax3.set_title('Two Days Prior Max Temp')

# Friend Estimate
ax4.plot(dates, features['friend'])
ax4.set_xlabel('Date'); ax4.set_ylabel('Temperature'); ax4.set_title('Friend Estimate')

plt.tight_layout(pad=2)


# # 数据预处理
# 

# ### One-Hot Encoding

# 原始数据：
# 
# | week |
# |------|
# | Mon  |
# | Tue  |
# | Wed  |
# | Thu  |
# | Fri  |
# 
# 编码转换后:
# 
# | Mon | Tue | Wed | Thu | Fri |
# |-----|-----|-----|-----|-----|
# | 1   | 0   | 0   | 0   | 0   |
# | 0   | 1   | 0   | 0   | 0   |
# | 0   | 0   | 1   | 0   | 0   |
# | 0   | 0   | 0   | 1   | 0   |
# | 0   | 0   | 0   | 0   | 1   |

# In[10]:


# One-hot encode categorical features
features = pd.get_dummies(features)
features.head(5)


# In[11]:


print('Shape of features after one-hot encoding:', features.shape)


# ### 标签与数据格式转换

# In[12]:


# Use numpy to convert to arrays
import numpy as np

# Labels are the values we want to predict
labels = np.array(features['actual'])

# Remove the labels from the features
# axis 1 refers to the columns
features= features.drop('actual', axis = 1)

# Saving feature names for later use
feature_list = list(features.columns)

# Convert to numpy array
features = np.array(features)


# ### 训练集与测试集

# In[13]:


# Using Skicit-learn to split data into training and testing sets
from sklearn.model_selection import train_test_split

# Split the data into training and testing sets
train_features, test_features, train_labels, test_labels = train_test_split(features, labels, test_size = 0.25,
                                                                           random_state = 42)


# In[14]:


print('Training Features Shape:', train_features.shape)
print('Training Labels Shape:', train_labels.shape)
print('Testing Features Shape:', test_features.shape)
print('Testing Labels Shape:', test_labels.shape)


# # 建立一个基础的随机森林模型

# In[16]:


# Import the model we are using
from sklearn.ensemble import RandomForestRegressor

# Instantiate model 
rf = RandomForestRegressor(n_estimators= 1000, random_state=42)

# Train the model on training data
rf.fit(train_features, train_labels);


# # 测试

# In[17]:


# Use the forest's predict method on the test data
predictions = rf.predict(test_features)

# Calculate the absolute errors
errors = abs(predictions - test_labels)

# Print out the mean absolute error (mae)
print('Mean Absolute Error:', round(np.mean(errors), 2), 'degrees.')


# # MAPE指标
# mape( mean absolute percent error)

# In[18]:


# Calculate mean absolute percentage error (MAPE)
mape = 100 * (errors / test_labels)

# Calculate and display accuracy
accuracy = 100 - np.mean(mape)
print('Accuracy:', round(accuracy, 2), '%.')


# # 可视化展示树

# In[23]:


rf_new = RandomForestRegressor(n_estimators = 100, criterion = 'mse', max_depth = None, 
                               min_samples_split = 2, min_samples_leaf = 1)


# # Interpret Model Results

# ## Visualizing a Single Decision Tree

# In[24]:


# Import tools needed for visualization
from sklearn.tree import export_graphviz
import pydot #pip install pydot

# Pull out one tree from the forest
tree = rf.estimators_[5]

# Export the image to a dot file
export_graphviz(tree, out_file = 'tree.dot', feature_names = feature_list, rounded = True, precision = 1)

# Use dot file to create a graph
(graph, ) = pydot.graph_from_dot_file('tree.dot')

# Write graph to a png file
graph.write_png('tree.png'); 


# ![Decision Tree](tree.png)

# In[25]:


print('The depth of this tree is:', tree.tree_.max_depth)


# 还是小一点吧。。。

# In[26]:


# Limit depth of tree to 2 levels
rf_small = RandomForestRegressor(n_estimators=10, max_depth = 3, random_state=42)
rf_small.fit(train_features, train_labels)

# Extract the small tree
tree_small = rf_small.estimators_[5]

# Save the tree as a png image
export_graphviz(tree_small, out_file = 'small_tree.dot', feature_names = feature_list, rounded = True, precision = 1)

(graph, ) = pydot.graph_from_dot_file('small_tree.dot')

graph.write_png('small_tree.png');


# ![Small Decision Tree](images/small_tree.PNG)

# ### Annotated Version of Tree

# ![Annotated Decision Tree](images/small_tree_annotated.PNG)

# ## 特征重要性

# In[27]:


# Get numerical feature importances
importances = list(rf.feature_importances_)

# List of tuples with variable and importance
feature_importances = [(feature, round(importance, 2)) for feature, importance in zip(feature_list, importances)]

# Sort the feature importances by most important first
feature_importances = sorted(feature_importances, key = lambda x: x[1], reverse = True)

# Print out the feature and importances 
[print('Variable: {:20} Importance: {}'.format(*pair)) for pair in feature_importances];


# ### 用最重要的特征再来试试

# In[28]:


# New random forest with only the two most important variables
rf_most_important = RandomForestRegressor(n_estimators= 1000, random_state=42)

# Extract the two most important features
important_indices = [feature_list.index('temp_1'), feature_list.index('average')]
train_important = train_features[:, important_indices]
test_important = test_features[:, important_indices]

# Train the random forest
rf_most_important.fit(train_important, train_labels)

# Make predictions and determine the error
predictions = rf_most_important.predict(test_important)

errors = abs(predictions - test_labels)

# Display the performance metrics
print('Mean Absolute Error:', round(np.mean(errors), 2), 'degrees.')

mape = np.mean(100 * (errors / test_labels))
accuracy = 100 - mape

print('Accuracy:', round(accuracy, 2), '%.')


# In[22]:


# list of x locations for plotting
x_values = list(range(len(importances)))

# Make a bar chart
plt.bar(x_values, importances, orientation = 'vertical')

# Tick labels for x axis
plt.xticks(x_values, feature_list, rotation='vertical')

# Axis labels and title
plt.ylabel('Importance'); plt.xlabel('Variable'); plt.title('Variable Importances'); 


# ### 预测值与真实值之间的差异

# In[23]:


# Dates of training values
months = features[:, feature_list.index('month')]
days = features[:, feature_list.index('day')]
years = features[:, feature_list.index('year')]

# List and then convert to datetime object
dates = [str(int(year)) + '-' + str(int(month)) + '-' + str(int(day)) for year, month, day in zip(years, months, days)]
dates = [datetime.datetime.strptime(date, '%Y-%m-%d') for date in dates]

# Dataframe with true values and dates
true_data = pd.DataFrame(data = {'date': dates, 'actual': labels})

# Dates of predictions
months = test_features[:, feature_list.index('month')]
days = test_features[:, feature_list.index('day')]
years = test_features[:, feature_list.index('year')]

# Column of dates
test_dates = [str(int(year)) + '-' + str(int(month)) + '-' + str(int(day)) for year, month, day in zip(years, months, days)]

# Convert to datetime objects
test_dates = [datetime.datetime.strptime(date, '%Y-%m-%d') for date in test_dates]

# Dataframe with predictions and dates
predictions_data = pd.DataFrame(data = {'date': test_dates, 'prediction': predictions}) 


# In[24]:


# Plot the actual values
plt.plot(true_data['date'], true_data['actual'], 'b-', label = 'actual')

# Plot the predicted values
plt.plot(predictions_data['date'], predictions_data['prediction'], 'ro', label = 'prediction')
plt.xticks(rotation = '60'); 
plt.legend()

# Graph labels
plt.xlabel('Date'); plt.ylabel('Maximum Temperature (F)'); plt.title('Actual and Predicted Values');