Методы машинного обучения
Семинар: метрические методы
In [57]:
%matplotlib inline
In [58]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
plt.style.use('ggplot')
plt.rcParams['figure.figsize'] = (12,8)
Стоимоcть жилья в Калифорнии¶
Будем использовать стандартный датасет из sklearn (Нужен интернет)
In [59]:
from sklearn.datasets import fetch_california_housing
In [60]:
data = fetch_california_housing()
X = data.data
y = data.target
In [61]:
X.shape
Out[61]:
(20640, 8)
In [62]:
plt.hist(y)
Out[62]:
(array([ 877., 3612., 4099., 3771., 2799., 1769., 1239., 752., 479.,
1243.]),
array([0.14999 , 0.634992, 1.119994, 1.604996, 2.089998, 2.575 ,
3.060002, 3.545004, 4.030006, 4.515008, 5.00001 ]),
<a list of 10 Patch objects>)
In [63]:
print(data.DESCR)
.. _california_housing_dataset:
California Housing dataset
--------------------------
**Data Set Characteristics:**
:Number of Instances: 20640
:Number of Attributes: 8 numeric, predictive attributes and the target
:Attribute Information:
- MedInc median income in block
- HouseAge median house age in block
- AveRooms average number of rooms
- AveBedrms average number of bedrooms
- Population block population
- AveOccup average house occupancy
- Latitude house block latitude
- Longitude house block longitude
:Missing Attribute Values: None
This dataset was obtained from the StatLib repository.
http://lib.stat.cmu.edu/datasets/
The target variable is the median house value for California districts.
This dataset was derived from the 1990 U.S. census, using one row per census
block group. A block group is the smallest geographical unit for which the U.S.
Census Bureau publishes sample data (a block group typically has a population
of 600 to 3,000 people).
It can be downloaded/loaded using the
:func:`sklearn.datasets.fetch_california_housing` function.
.. topic:: References
- Pace, R. Kelley and Ronald Barry, Sparse Spatial Autoregressions,
Statistics and Probability Letters, 33 (1997) 291-297
- Нанесите координаты домов и раскрасте их в зависимости от их стоимости
- HINT:
plt.scatter()
- HINT:
In [64]:
plt.scatter(X[:, -1], X[:, -2], c=y)
Out[64]:
<matplotlib.collections.PathCollection at 0x1a1e7e76a0>
- Рассмотрим только координаты домов в качестве признаков. В качестве меры близости будем использовать расстояние на сфере (Земле) между координатами. Для этого, для начала, надо перевести координаты из градусов в радианы
- HINT:
np.radians
- HINT:
- Расделите данные на обучающую и контрольную выборку
- HINT:
sklearn.model_selection.train_test_split()и не забудте указатьrandom_state
- HINT:
In [65]:
X = X[:, -2:]
In [66]:
X = np.radians(X)
In [67]:
from sklearn.model_selection import train_test_split
In [68]:
X_train, X_valid, y_train, y_valid = train_test_split(X, y, test_size=0.2, random_state=123)
In [69]:
X_train.shape
Out[69]:
(16512, 2)
In [70]:
X_valid.shape
Out[70]:
(4128, 2)
- Создайте kNN классификатор с метрикой
haversine- HINT: sklearn.metrics
- Для k = 1..15 изобразите среднее абсолютное отклонение ответов модели на обучающей и контрольной выборке
- Какие методологические ошибки были допущены
- Какое значение $k$ оптимально?
In [71]:
from sklearn.neighbors import KNeighborsRegressor
In [72]:
from sklearn.metrics import mean_absolute_error
In [73]:
knn = KNeighborsRegressor(n_neighbors=10, metric='haversine')
In [74]:
knn.fit(X_train, y_train)
Out[74]:
KNeighborsRegressor(algorithm='auto', leaf_size=30, metric='haversine',
metric_params=None, n_jobs=None, n_neighbors=10, p=2,
weights='uniform')
In [75]:
y_hat = knn.predict(X_valid)
In [76]:
y_hat.shape
Out[76]:
(4128,)
In [77]:
mean_absolute_error(y_hat, y_valid)
Out[77]:
0.33340470712209297
In [78]:
quality_valid = []
quality_train = []
for k in range(1,16):
knn = KNeighborsRegressor(n_neighbors=k, metric='haversine')
knn.fit(X_train, y_train)
y_hat_valid = knn.predict(X_valid)
y_hat_train = knn.predict(X_train)
quality_valid.append(mean_absolute_error(y_valid, y_hat_valid))
quality_train.append(mean_absolute_error(y_train, y_hat_train))
In [79]:
k_range = range(1,16)
plt.plot(k_range, quality_valid, color='blue', label='valid')
plt.plot(k_range, quality_train, label='train')
plt.xlabel('number of neighbours')
plt.ylabel('MAE quality')
plt.legend()
Out[79]:
<matplotlib.legend.Legend at 0x1a1cf83748>
Оценка уровня заработка¶
1. Categorical Attributes¶
- workclass: (categorical) Private, Self-emp-not-inc, Self-emp-inc, Federal-gov, Local-gov, State-gov, Without-pay, Never-worked.
- Individual work category
- education: (categorical) Bachelors, Some-college, 11th, HS-grad, Prof-school, Assoc-acdm, Assoc-voc, 9th, 7th-8th, 12th, Masters, 1st-4th, 10th, Doctorate, 5th-6th, Preschool.
- Individual's highest education degree
- marital-status: (categorical) Married-civ-spouse, Divorced, Never-married, Separated, Widowed, Married-spouse-absent, Married-AF-spouse.
- Individual marital status
- occupation: (categorical) Tech-support, Craft-repair, Other-service, Sales, Exec-managerial, Prof-specialty, Handlers-cleaners, Machine-op-inspct, Adm-clerical, Farming-fishing, Transport-moving, Priv-house-serv, Protective-serv, Armed-Forces.
- Individual's occupation
- relationship: (categorical) Wife, Own-child, Husband, Not-in-family, Other-relative, Unmarried.
- Individual's relation in a family
- race: (categorical) White, Asian-Pac-Islander, Amer-Indian-Eskimo, Other, Black.
- Race of Individual
- sex: (categorical) Female, Male.
- native-country: (categorical) United-States, Cambodia, England, Puerto-Rico, Canada, Germany, Outlying-US(Guam-USVI-etc), India, Japan, Greece, South, China, Cuba, Iran, Honduras, Philippines, Italy, Poland, Jamaica, Vietnam, Mexico, Portugal, Ireland, France, Dominican-Republic, Laos, Ecuador, Taiwan, Haiti, Columbia, Hungary, Guatemala, Nicaragua, Scotland, Thailand, Yugoslavia, El-Salvador, Trinadad&Tobago, Peru, Hong, Holand-Netherlands.
- Individual's native country
2. Continuous Attributes¶
- age: continuous.
- Age of an individual
- education-num: number of education year, continuous.
- Individual's year of receiving education
- fnlwgt: final weight, continuous.
- The weights on the CPS files are controlled to independent estimates of the civilian noninstitutional population of the US. These are prepared monthly for us by Population Division here at the Census Bureau.
- capital-gain: continuous.
- capital-loss: continuous.
- hours-per-week: continuous.
- Individual's working hour per week
In [80]:
df = pd.read_csv('./data/adult.csv.gz')
In [81]:
df.head()
Out[81]:
| age | workclass | fnlwgt | education | educational-num | marital-status | occupation | relationship | race | gender | capital-gain | capital-loss | hours-per-week | native-country | income | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 25 | Private | 226802 | 11th | 7 | Never-married | Machine-op-inspct | Own-child | Black | Male | 0 | 0 | 40 | United-States | <=50K |
| 1 | 38 | Private | 89814 | HS-grad | 9 | Married-civ-spouse | Farming-fishing | Husband | White | Male | 0 | 0 | 50 | United-States | <=50K |
| 2 | 28 | Local-gov | 336951 | Assoc-acdm | 12 | Married-civ-spouse | Protective-serv | Husband | White | Male | 0 | 0 | 40 | United-States | >50K |
| 3 | 44 | Private | 160323 | Some-college | 10 | Married-civ-spouse | Machine-op-inspct | Husband | Black | Male | 7688 | 0 | 40 | United-States | >50K |
| 4 | 18 | ? | 103497 | Some-college | 10 | Never-married | ? | Own-child | White | Female | 0 | 0 | 30 | United-States | <=50K |
- Проанализируйте таблицу на предмет пропущенных значений: сколько их, в каких столбцах их много?
In [82]:
df.isnull().mean()
# Напрямую пропуски не выявляются. Вызывает подозрение знак "?" Давайте скажем pandas, что "?" - это пропуск
Out[82]:
age 0.0 workclass 0.0 fnlwgt 0.0 education 0.0 educational-num 0.0 marital-status 0.0 occupation 0.0 relationship 0.0 race 0.0 gender 0.0 capital-gain 0.0 capital-loss 0.0 hours-per-week 0.0 native-country 0.0 income 0.0 dtype: float64
In [83]:
df = pd.read_csv('./data/adult.csv.gz', na_values=['?'])
df.isnull().mean()
Out[83]:
age 0.000000 workclass 0.057307 fnlwgt 0.000000 education 0.000000 educational-num 0.000000 marital-status 0.000000 occupation 0.057512 relationship 0.000000 race 0.000000 gender 0.000000 capital-gain 0.000000 capital-loss 0.000000 hours-per-week 0.000000 native-country 0.017546 income 0.000000 dtype: float64
In [84]:
df = df.dropna() # удалим пропуски
- Рассмотрите категориальные переменные. Что можно сделать с редкими категориями?
- Выполните переобозначение значений категориальных признаков (где необходимо)
In [85]:
categ_cols = ['education', 'marital-status', 'workclass', 'occupation', 'relationship', 'race', 'gender', 'native-country']
for col in categ_cols:
print('='*10)
print(col)
print(df.loc[:, col].value_counts())
========== education HS-grad 14783 Some-college 9899 Bachelors 7570 Masters 2514 Assoc-voc 1959 11th 1619 Assoc-acdm 1507 10th 1223 7th-8th 823 Prof-school 785 9th 676 12th 577 Doctorate 544 5th-6th 449 1st-4th 222 Preschool 72 Name: education, dtype: int64 ========== marital-status Married-civ-spouse 21055 Never-married 14598 Divorced 6297 Separated 1411 Widowed 1277 Married-spouse-absent 552 Married-AF-spouse 32 Name: marital-status, dtype: int64 ========== workclass Private 33307 Self-emp-not-inc 3796 Local-gov 3100 State-gov 1946 Self-emp-inc 1646 Federal-gov 1406 Without-pay 21 Name: workclass, dtype: int64 ========== occupation Craft-repair 6020 Prof-specialty 6008 Exec-managerial 5984 Adm-clerical 5540 Sales 5408 Other-service 4808 Machine-op-inspct 2970 Transport-moving 2316 Handlers-cleaners 2046 Farming-fishing 1480 Tech-support 1420 Protective-serv 976 Priv-house-serv 232 Armed-Forces 14 Name: occupation, dtype: int64 ========== relationship Husband 18666 Not-in-family 11702 Own-child 6626 Unmarried 4788 Wife 2091 Other-relative 1349 Name: relationship, dtype: int64 ========== race White 38903 Black 4228 Asian-Pac-Islander 1303 Amer-Indian-Eskimo 435 Other 353 Name: race, dtype: int64 ========== gender Male 30527 Female 14695 Name: gender, dtype: int64 ========== native-country United-States 41292 Mexico 903 Philippines 283 Germany 193 Puerto-Rico 175 Canada 163 El-Salvador 147 India 147 Cuba 133 England 119 China 113 Jamaica 103 South 101 Italy 100 Dominican-Republic 97 Japan 89 Guatemala 86 Vietnam 83 Columbia 82 Poland 81 Haiti 69 Portugal 62 Iran 56 Taiwan 55 Greece 49 Nicaragua 48 Peru 45 Ecuador 43 Ireland 36 France 36 Thailand 29 Hong 28 Trinadad&Tobago 26 Cambodia 26 Yugoslavia 23 Outlying-US(Guam-USVI-etc) 22 Laos 21 Scotland 20 Honduras 19 Hungary 18 Holand-Netherlands 1 Name: native-country, dtype: int64
- Рассмотрите попарные зависимости между признаками
- Какие есть взаимосвязи с целевым признаком
In [86]:
# Во-первых можно понять, что признаки с постфиксом "-num" кодируют категории
pd.crosstab(df.education, df.loc[:, 'educational-num'])
Out[86]:
| educational-num | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| education | ||||||||||||||||
| 10th | 0 | 0 | 0 | 0 | 0 | 1223 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 11th | 0 | 0 | 0 | 0 | 0 | 0 | 1619 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 12th | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 577 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 1st-4th | 0 | 222 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 5th-6th | 0 | 0 | 449 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 7th-8th | 0 | 0 | 0 | 823 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 9th | 0 | 0 | 0 | 0 | 676 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Assoc-acdm | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1507 | 0 | 0 | 0 | 0 |
| Assoc-voc | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1959 | 0 | 0 | 0 | 0 | 0 |
| Bachelors | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 7570 | 0 | 0 | 0 |
| Doctorate | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 544 |
| HS-grad | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 14783 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Masters | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 2514 | 0 | 0 |
| Preschool | 72 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Prof-school | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 785 | 0 |
| Some-college | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 9899 | 0 | 0 | 0 | 0 | 0 | 0 |
In [87]:
df.corr(method='spearman') # Корреляция между признаками везде слабая
Out[87]:
| age | fnlwgt | educational-num | capital-gain | capital-loss | hours-per-week | |
|---|---|---|---|---|---|---|
| age | 1.000000 | -0.077455 | 0.067171 | 0.121161 | 0.060635 | 0.157484 |
| fnlwgt | -0.077455 | 1.000000 | -0.032139 | -0.009988 | -0.002111 | -0.020791 |
| educational-num | 0.067171 | -0.032139 | 1.000000 | 0.119901 | 0.077575 | 0.165424 |
| capital-gain | 0.121161 | -0.009988 | 0.119901 | 1.000000 | -0.067296 | 0.093086 |
| capital-loss | 0.060635 | -0.002111 | 0.077575 | -0.067296 | 1.000000 | 0.060391 |
| hours-per-week | 0.157484 | -0.020791 | 0.165424 | 0.093086 | 0.060391 | 1.000000 |
In [88]:
df.head()
Out[88]:
| age | workclass | fnlwgt | education | educational-num | marital-status | occupation | relationship | race | gender | capital-gain | capital-loss | hours-per-week | native-country | income | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 25 | Private | 226802 | 11th | 7 | Never-married | Machine-op-inspct | Own-child | Black | Male | 0 | 0 | 40 | United-States | <=50K |
| 1 | 38 | Private | 89814 | HS-grad | 9 | Married-civ-spouse | Farming-fishing | Husband | White | Male | 0 | 0 | 50 | United-States | <=50K |
| 2 | 28 | Local-gov | 336951 | Assoc-acdm | 12 | Married-civ-spouse | Protective-serv | Husband | White | Male | 0 | 0 | 40 | United-States | >50K |
| 3 | 44 | Private | 160323 | Some-college | 10 | Married-civ-spouse | Machine-op-inspct | Husband | Black | Male | 7688 | 0 | 40 | United-States | >50K |
| 5 | 34 | Private | 198693 | 10th | 6 | Never-married | Other-service | Not-in-family | White | Male | 0 | 0 | 30 | United-States | <=50K |
In [89]:
# По некоторым признакам прослеживается четкое различие между классами
numeric_features = ['educational-num', 'age', 'fnlwgt', 'capital-gain', 'capital-loss', 'hours-per-week']
for col in numeric_features:
df.boxplot(column=col, by='income')
In [90]:
pd.crosstab(df.income, df.gender, normalize=True, margins=True)
Out[90]:
| gender | Female | Male | All |
|---|---|---|---|
| income | |||
| <=50K | 0.288046 | 0.464110 | 0.752156 |
| >50K | 0.036907 | 0.210937 | 0.247844 |
| All | 0.324952 | 0.675048 | 1.000000 |
In [91]:
pd.crosstab(df.income, df.race, normalize=True, margins=True)
Out[91]:
| race | Amer-Indian-Eskimo | Asian-Pac-Islander | Black | Other | White | All |
|---|---|---|---|---|---|---|
| income | ||||||
| <=50K | 0.008447 | 0.020654 | 0.081686 | 0.006811 | 0.634558 | 0.752156 |
| >50K | 0.001172 | 0.008160 | 0.011808 | 0.000995 | 0.225709 | 0.247844 |
| All | 0.009619 | 0.028813 | 0.093494 | 0.007806 | 0.860267 | 1.000000 |
- Преобразуйте данные таким образом, чтобы их можно было подать на вход в классификатор (никаких строк, категориальные признаки в OneHotEncoding)
In [92]:
df.head()
Out[92]:
| age | workclass | fnlwgt | education | educational-num | marital-status | occupation | relationship | race | gender | capital-gain | capital-loss | hours-per-week | native-country | income | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 25 | Private | 226802 | 11th | 7 | Never-married | Machine-op-inspct | Own-child | Black | Male | 0 | 0 | 40 | United-States | <=50K |
| 1 | 38 | Private | 89814 | HS-grad | 9 | Married-civ-spouse | Farming-fishing | Husband | White | Male | 0 | 0 | 50 | United-States | <=50K |
| 2 | 28 | Local-gov | 336951 | Assoc-acdm | 12 | Married-civ-spouse | Protective-serv | Husband | White | Male | 0 | 0 | 40 | United-States | >50K |
| 3 | 44 | Private | 160323 | Some-college | 10 | Married-civ-spouse | Machine-op-inspct | Husband | Black | Male | 7688 | 0 | 40 | United-States | >50K |
| 5 | 34 | Private | 198693 | 10th | 6 | Never-married | Other-service | Not-in-family | White | Male | 0 | 0 | 30 | United-States | <=50K |
In [93]:
drop_cols = ['education', 'native-country', 'relationship']
df = df.drop(drop_cols, axis=1)
idx = df.workclass == 'Without-pay'
df = df.loc[~idx, :]
cols_replaces = {'occupation': {'Armed-Forces': 'Protective-serv'},
'gender': {'Male': 1, 'Female': 0},
'marital-status': {'Married-civ-spouse': 'Married',
'Married-spouse-absent': 'Married',
'Married-AF-spouse': 'Married',
'Separated': 'Alone',
'Divorced': 'Alone',
'Never-married ': 'Alone',
'Widowed': 'Alone'},
'income': {'<=50K': 0, '>50K': 1}}
df = df.replace(cols_replaces)
df = pd.get_dummies(df, columns=['workclass', 'marital-status', 'occupation', 'race'], prefix_sep='=')
In [94]:
df.shape
Out[94]:
(45201, 35)
- Разделите данные на обучающую и валидационную выборку
- Повторите упражнение с выбором значения параметра $k$
In [95]:
label_col = df.columns == 'income'
X = df.loc[:, ~label_col].values
y = df.loc[:, label_col].values.flatten()
In [99]:
from sklearn.model_selection import train_test_split
from sklearn.metrics import roc_auc_score
from sklearn.neighbors import KNeighborsClassifier
Сделаем не через train-test split, а через кросс-валидацию. Посмотрим, что делает функция validation_curve.
In [111]:
from sklearn.model_selection import validation_curve
from sklearn.model_selection import StratifiedKFold
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import MinMaxScaler
validation_curve позволяе исследовать влияние отдельного гиперпараметра модели на ее качество.
Построим валидационные кривые по гиперпараметру глубина дерева (max_depth) остальных гиперпараметрах, зафиксированных на значениях по-умолчанию
In [116]:
## Обычный KNN
# model = KNeighborsClassifier()
## KNN с нормализацией признаков
model = Pipeline([
('scaler', MinMaxScaler()),
('knn', KNeighborsClassifier())
])
cv = StratifiedKFold(n_splits=5, shuffle=True, random_state=132)
k = range(1, 15)
knn_train_scores, knn_valid_scores = validation_curve(model, X, y,
'knn__n_neighbors', k,
cv=cv, scoring='roc_auc')
# это значит, что мы будем проверять влияние параметра n_neighbors
# в дапозоне от 1 до 15
# и для этого мы будем использовать 5-fold кросс-валидацию
# с мерой качества roc_auc
/Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning) /Users/andrey.shestakov/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:590: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler. warnings.warn(msg, DataConversionWarning)
In [117]:
knn_train_scores.shape
Out[117]:
(14, 5)
In [118]:
knn_valid_scores.shape
Out[118]:
(14, 5)
Посмотрите, что изображает синяя и красная линия
In [120]:
train_score_mean = knn_train_scores.mean(axis=1)
train_score_std = knn_train_scores.std(axis=1)
valid_scores_mean = knn_valid_scores.mean(axis=1)
valid_scores_std = knn_valid_scores.std(axis=1)
plt.fill_between(k, train_score_mean-train_score_std, train_score_mean+train_score_std, color='b',
interpolate=True, alpha=0.5,)
plt.fill_between(k, valid_scores_mean-valid_scores_std, valid_scores_mean+valid_scores_std, color='r',
interpolate=True, alpha=0.5)
plt.plot(k, train_score_mean, c='b', lw=2, label='train')
plt.plot(k, valid_scores_mean, c='r', lw=2, label='validation')
plt.xlabel('k')
plt.ylabel('ROC AUC')
Out[120]:
Text(0,0.5,'ROC AUC')
In [ ]: