《Python数据科学手册》异常校正

Posted 2021-03-26 云忘归

由于一些模块的变迁，导致复现《python数据科学手册》代码（尤其第5章-机器学习）时，经常报错。

以下是我个人的一些校证。

如果诸位在学习《python数据科学手册》的过程中，遇到什么疑难，欢迎留言。
1. scikit-learn.cross_validation 模块变迁
自 `scikit-learn 0.20 `版起，已经用`model_selection`模块代替`cross_validation`模块。因此，复现代码时，`from sklearn.cross_validation import xxx` 时，会报出`ModuleNotFoundError: No module named ‘sklearn.cross_validation‘`的错误。P307:

In[15]:from sklearn.cross_validation import train_test_split # Error
In[15]:from sklearn.model_selection import train_test_split # Amend

In[20]:
from sklearn.mixture import GMM # Error
from sklearn.mixture import GaussianMixture # Amend

In[5]: from sklearn.cross_validation import train_test_split # Error
In[5]: from sklearn.model_selection import train_test_split # Amend


# 用 model_selection 替换 cross_validation
In[7]: from sklearn.cross_validation import cross_val_scroe # Error
In[7]: from sklearn.model_selection import cross_val_scroe # Amend


In[8]: from sklearn.cross_validation import cross_val_scroe # Error
scores = cross_val_score(model, X, y, cv=LeaveOneOut(len(X)) # Error
In[8]: from sklearn.model_selection import cross_val_scroe # Amend
scores = cross_val_score(model, X, y, cv=LeaveOneOut() # Amend，去掉 len(X)

 

2. scikit-learn.learning_curve 模块变迁
自 `scikit-learn 0.20 `版起，已经用`model_selection`模块代替`learning_curve`模块。因此，复现代码时，`from sklearn.learning_curve import xxx` 时，会报出`ModuleNotFoundError: No module named ‘sklearn.learning_curve‘`的错误。

P321:
In[13]:
from sklearn.learning_curve import validation_curve  # Error
from sklearn.model_selection import validation_curve  # Amend


P325:
In[17]:
from sklearn.learning_curve import learning_curve  # Error
from sklearn.model_selection import learning_curve  # Amend

3. scikit-learn.grid_search 模块变迁

自 `scikit-learn 0.20 `版起，已经用`model_selection`模块代替`grid_search`模块。因此，复现代码时，`from sklearn.grid_search import xxx` 时，会报出`ModuleNotFoundError: No module named ‘sklearn.grid_search‘`的错误。

P326

In[18]:from sklearn.grid_search import GridSearchCV # Error
In[18]:from sklearn.model_selection import GridSearchCV # Amend

In[21]: plt.plot(X_test.ravel(), y_test, hold=True); # Error
In[21]: plt.plot(X_test.ravel(), y_test); # Amend, 去掉 hold=True

 

4. 其他错误

P248:
In[3]:
ax = plt.axes(axisbg=‘#E6E6E6‘) # Error
ax = plt.axes(facecolor=‘#E6E6E6‘) # Amend, axisbg -> facecolor


P275:
In[6]:
plt.hist(data[col], normed=True, alpha=0.5) # Error
plt.hist(data[col], density=True, alpha=0.5) # Amend, normed -> density


P248:
In[3]:
ax = plt.axes(axisbg=‘#E6E6E6‘) # Error
ax = plt.axes(facecolor=‘#E6E6E6‘) # Amend, axisbg -> facecolor


P275:
In[6]:
plt.hist(data[col], normed=True, alpha=0.5) # Error
plt.hist(data[col], density=True, alpha=0.5) # Amend, normed -> density


P279:
In[13]:
sns.pairplot(iris, hue=‘species‘, size=2.5) # Error
sns.pairplot(iris, hue=‘species‘, height=2.5) # Amend, size -> height


P301:
In[2]:
sns.parirplot(iris, hue=‘species‘, size=1.5); # Error
sns.parirplot(iris, hue=‘species‘, height=1.5); # Amend, size 改为 height


P349:
In[14]:
weather = pd.read_csv(‘599021.csv‘, index_col=‘DATE‘, parse_dates=True) # Error
weather = pd.read_csv(‘599021.csv‘, index_col=‘DATE‘, parse_dates=True) # Amend, 599021.csv -> BicycleWeather.csv

In[15]: daily = counts.resample(‘d‘, how=‘sum‘) # Error
In[15]: daily = counts.resample(‘d‘).sum() # Amend


P361:
In[14]: clf = SVC(kernel=‘rbf‘, C=1E6) # Error
In[14]: clf = SVC(kernel=‘rbf‘, C=1E6, gamma=‘auto‘) # Amend, add gamma=‘auto‘


P363
In[20]:
from sklearn.decomposition import RandomizedPCA # Error
pac = RandomizedPCA(n_components=150, whiten=True, random_state=42) # Error

from sklearn.decomposition import PCA # Amend, RandomizedPCA -> PCA
pac = PCA(n_components=150, whiten=True, random_state=42) # Amend, RandomizedPCA -> PCA


P364:
In[21]: from sklearn.cross_validation import train_test_split # Error
In[21]: from sklearn.model_selection import train_test_split # Amend

In[22]: from sklearn.grid_search import GridSearchCV # Error
grid = GridSearchCV(model, param_grid) # Error
In[22]: from sklearn.model_selection import GridSearchCV # Amend
grid = GridSearchCV(model, param_grid, cv=3) # Amend, add cv=3


P279:
In[13]:
sns.pairplot(iris, hue=‘species‘, size=2.5) # Error
sns.pairplot(iris, hue=‘species‘, height=2.5) # Amend, size -> height


P301:
In[2]:
sns.parirplot(iris, hue=‘species‘, size=1.5); # Error
sns.parirplot(iris, hue=‘species‘, height=1.5); # Amend, size 改为 height


P349:
In[14]:
weather = pd.read_csv(‘599021.csv‘, index_col=‘DATE‘, parse_dates=True) # Error
weather = pd.read_csv(‘599021.csv‘, index_col=‘DATE‘, parse_dates=True) # Amend, 599021.csv -> BicycleWeather.csv

In[15]: daily = counts.resample(‘d‘, how=‘sum‘) # Error
In[15]: daily = counts.resample(‘d‘).sum() # Amend


P361:
In[14]: clf = SVC(kernel=‘rbf‘, C=1E6) # Error
In[14]: clf = SVC(kernel=‘rbf‘, C=1E6, gamma=‘auto‘) # Amend, add gamma=‘auto‘


P248:
In[3]:
ax = plt.axes(axisbg=‘#E6E6E6‘) # Error
ax = plt.axes(facecolor=‘#E6E6E6‘) # Amend, axisbg -> facecolor


P275:
In[6]:
plt.hist(data[col], normed=True, alpha=0.5) # Error
plt.hist(data[col], density=True, alpha=0.5) # Amend, normed -> density


P279:
In[13]:
sns.pairplot(iris, hue=‘species‘, size=2.5) # Error
sns.pairplot(iris, hue=‘species‘, height=2.5) # Amend, size -> height


P301:
In[2]:
sns.parirplot(iris, hue=‘species‘, size=1.5); # Error
sns.parirplot(iris, hue=‘species‘, height=1.5); # Amend, size 改为 height


P349:
In[14]:
weather = pd.read_csv(‘599021.csv‘, index_col=‘DATE‘, parse_dates=True) # Error
weather = pd.read_csv(‘599021.csv‘, index_col=‘DATE‘, parse_dates=True) # Amend, 599021.csv -> BicycleWeather.csv

In[15]: daily = counts.resample(‘d‘, how=‘sum‘) # Error
In[15]: daily = counts.resample(‘d‘).sum() # Amend


P361:
In[14]: clf = SVC(kernel=‘rbf‘, C=1E6) # Error
In[14]: clf = SVC(kernel=‘rbf‘, C=1E6, gamma=‘auto‘) # Amend, add gamma=‘auto‘


P363:
In[20]:
from sklearn.decomposition import RandomizedPCA # Error
pac = RandomizedPCA(n_components=150, whiten=True, random_state=42) # Error

from sklearn.decomposition import PCA # Amend, RandomizedPCA -> PCA
pac = PCA(n_components=150, whiten=True, random_state=42) # Amend, RandomizedPCA -> PCA


P364:
In[21]: from sklearn.cross_validation import train_test_split # Error
In[21]: from sklearn.model_selection import train_test_split # Amend

In[22]: from sklearn.grid_search import GridSearchCV # Error
grid = GridSearchCV(model, param_grid) # Error
In[22]: from sklearn.model_selection import GridSearchCV # Amend
grid = GridSearchCV(model, param_grid, cv=3) # Amend, add cv=3

P363
In[20]:
from sklearn.decomposition import RandomizedPCA # Error
pac = RandomizedPCA(n_components=150, whiten=True, random_state=42) # Error

from sklearn.decomposition import PCA # Amend, RandomizedPCA -> PCA
pac = PCA(n_components=150, whiten=True, random_state=42) # Amend, RandomizedPCA -> PCA

P364:
In[21]: from sklearn.cross_validation import train_test_split # Error
In[21]: from sklearn.model_selection import train_test_split # Amend

In[22]: from sklearn.grid_search import GridSearchCV # Error
grid = GridSearchCV(model, param_grid) # Error
In[22]: from sklearn.model_selection import GridSearchCV # Amend
grid = GridSearchCV(model, param_grid, cv=3) # Amend, add cv=3