> For the complete documentation index, see [llms.txt](https://jen-hsuan-hsieh.gitbook.io/python/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://jen-hsuan-hsieh.gitbook.io/python/chapter-2courses/21python-for-data-science-and-machine-learning-bootcamp/sd/21132knn-with-python.md).

# 2.1.13.2.KNN with Python

## 1. 匯入基本的library

* [pandas](https://jenhsuan.gitbooks.io/python/content/chapter-2courses/21python-for-data-science-and-machine-learning-bootcamp/211jupyter-overview/214python-for-data-analysis-pandas.html), [numpy](https://jenhsuan.gitbooks.io/python/content/chapter-2courses/21python-for-data-science-and-machine-learning-bootcamp/211jupyter-overview/213python-for-data-analysis-numpy.html), [matplotlib](https://jenhsuan.gitbooks.io/python/content/test.html), [seaborn](https://jenhsuan.gitbooks.io/python/content/217python-for-data-visualization-seaborn.html)

  ```
    import pandas as pd
    import numpy as np
    import matplotlib.pyplot as plt
    import seaborn as sns
  ```
* 將圖表直接嵌入到Notebook之中

  ```
    %matplotlib inline
  ```

## 2.讀取資料並了解資料

```
df = pd.read_csv('Classified Data', index_col = 0)
df.head()
```

![](https://github.com/jenhsuan/python/tree/8fc9c0b8df4ccd709d3078c2d8842af0932de09d/assets/螢幕快照%202018-06-26%20下午10.46.37.png)

## 3.標準化資料

* 變數的scale通常對結果有很大的影響, 當使用KNN classifier時通常會統一observation的尺度
* 使用StandardScaler
  * [StandardScaler](https://www.cnblogs.com/chaosimple/p/4153167.html)的作用是將數據減去平均值並除以方差, 公式為(X-mean)/std
  * 將原始資料的TARGET CLASS drop掉後[fit, 再transform](https://blog.csdn.net/lz_peter/article/details/78237094)

    ```
    from sklearn.preprocessing import StandardScaler
    scaler  = StandardScaler()
    scaler.fit(df.drop("TARGET CLASS", axis = 1))
    scaled_features = scaler.transform(df.drop("TARGET CLASS", axis = 1))
    ```
* 將標準化後的資料準轉成DataFrame

  ```
    df_feat = pd.DataFrame(scaled_features, columns = df.columns[:-1])
    df_feat.head()
  ```

## 4.使用Skikit-learn library

* 首先介紹train\_test\_split, 這個函式可以隨機劃分訓練集和測試集

  ```
    from sklearn.cross_validation import train_test_split
    X = df_feat
    Y = df["TARGET CLASS"]
    X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.4, random_state=101)
  ```

## 5.使用KNN classifier進行預測

* 設定n\_neighbors(K)為1

  ```
    from sklearn.neighbors import KNeighborsClassifier
    knn = KNeighborsClassifier(n_neighbors = 1)
    knn.fit(X_train, y_train)
    predictions = knn.predict(X_test)
  ```

## 6.評估模型的精度

* 使用classification\_report

  ```
    from sklearn.metrics import classification_report
    print(classification_report(y_test, predictions))
  ```

  ![](https://github.com/jenhsuan/python/tree/8fc9c0b8df4ccd709d3078c2d8842af0932de09d/assets/螢幕快照%202018-06-26%20下午10.43.33.png)
* confusion\_matrix

  ```
    from sklearn.metrics import confusion_matrix
    print(confusion_matrix(y_test, predictions))
  ```

  ![](https://github.com/jenhsuan/python/tree/8fc9c0b8df4ccd709d3078c2d8842af0932de09d/assets/螢幕快照%202018-06-26%20下午10.45.32.png)

## 7.調整K值

* [將多次的預測值與y\_test的差異矩陣取平均](https://stackoverflow.com/questions/41419864/understand-np-mean-in-python)

  ```
    error_rate = []
    for i in range(1, 40):
        knn = KNeighborsClassifier(n_neighbors = i)
        knn.fit(X_train, y_train)
        pred_i = knn.predict(X_test)
        error_rate.append(np.mean(pred_i != y_test))

    plt.figure(figsize = (10, 6))
    plt.plot(range(1,40), error_rate, color = 'blue',linestyle = 'dashed', marker = 'o', markerfacecolor = 'red'
         , markersize = 10)
    plt.title("Error Rate vs K value")
    plt.xlabel('K')
  ```

  ![](https://github.com/jenhsuan/python/tree/8fc9c0b8df4ccd709d3078c2d8842af0932de09d/assets/螢幕快照%202018-06-26%20下午10.50.09.png)
* 重新預估k值

  ```
    knn = KNeighborsClassifier(n_neighbors = 37)
    knn.fit(X_train, y_train)
    predictions = knn.predict(X_test)
    print(classification_report(y_test, predictions))
    print(confusion_matrix(y_test, predictions))
  ```