Diamonds - Séquence de Préprocessing Simple (Encoding + Normalisation)

df = sns.load_dataset('diamonds')
df.head()

   carat      cut color clarity  depth  table  price     x     y     z
0   0.23    Ideal     E     SI2   61.5   55.0    326  3.95  3.98  2.43
1   0.21  Premium     E     SI1   59.8   61.0    326  3.89  3.84  2.31
2   0.23     Good     E     VS1   56.9   65.0    327  4.05  4.07  2.31
3   0.29  Premium     I     VS2   62.4   58.0    334  4.20  4.23  2.63
4   0.31     Good     J     SI2   63.3   58.0    335  4.34  4.35  2.75

from sklearn.model_selection import train_test_split
train_set, test_set = train_test_split(df, test_size=0.2, random_state=0)
train_set.shape

(43152, 10)

test_set.shape

(10788, 10)

cut_order = ['Fair', 'Good', 'Very Good', 'Premium', 'Ideal']
clarity_order = ['I1', 'SI2', 'SI1', 'VS2', 'VS1', 'VVS2', 'VVS1', 'IF']
color_order = ['E', 'I', 'J', 'H', 'F', 'G', 'D']

encoder = OrdinalEncoder(categories=[cut_order, color_order, clarity_order])
encoding_results = encoder.fit_transform(train_set[['cut', 'color', 'clarity']])

[[4. 5. 4.]
 [4. 5. 3.]
 [4. 0. 5.]
 ...
 [3. 1. 4.]
 [4. 5. 7.]
 [4. 4. 1.]]

train_set.loc[:, ['cut', 'color', 'clarity']] = encoding_results

       carat  cut  color  clarity  depth  table  price     x     y     z
26250   1.63  4.0    5.0      4.0   61.7   55.0  15697  7.56  7.60  4.68
31510   0.34  4.0    5.0      3.0   62.2   57.0    765  4.47  4.44  2.77
40698   0.40  4.0    0.0      5.0   61.7   56.0   1158  4.73  4.77  2.93
42634   0.58  3.0    3.0      2.0   62.1   55.0   1332  5.38  5.35  3.33
47714   0.63  2.0    6.0      2.0   62.8   57.0   1885  5.40  5.46  3.41
...      ...  ...    ...      ...    ...    ...    ...   ...   ...   ...
45891   0.52  3.0    4.0      3.0   60.7   59.0   1720  5.18  5.14  3.13
52416   0.70  1.0    6.0      2.0   63.6   60.0   2512  5.59  5.51  3.51
42613   0.32  3.0    1.0      4.0   61.3   58.0    505  4.35  4.39  2.68
43567   0.41  4.0    5.0      7.0   61.0   57.0   1431  4.81  4.79  2.93
2732    0.91  4.0    4.0      1.0   61.1   55.0   3246  6.24  6.19  3.80

[43152 rows x 10 columns]

normaliser = MinMaxScaler()
train_set = normaliser.fit_transform(train_set)

[[0.2972973  1.         0.83333333 ... 0.70391061 0.12903226 0.14716981]
 [0.02910603 1.         0.83333333 ... 0.41620112 0.075382   0.08710692]
 [0.04158004 1.         0.         ... 0.44040968 0.08098472 0.09213836]
 ...
 [0.02494802 0.75       0.16666667 ... 0.40502793 0.07453311 0.08427673]
 [0.04365904 1.         0.83333333 ... 0.44785847 0.08132428 0.09213836]
 [0.14760915 1.         0.66666667 ... 0.58100559 0.10509338 0.11949686]]

pd.DataFrame(train_set, columns=df.columns)

          carat   cut     color  ...         x         y         z
0      0.297297  1.00  0.833333  ...  0.703911  0.129032  0.147170
1      0.029106  1.00  0.833333  ...  0.416201  0.075382  0.087107
2      0.041580  1.00  0.000000  ...  0.440410  0.080985  0.092138
3      0.079002  0.75  0.500000  ...  0.500931  0.090832  0.104717
4      0.089397  0.50  1.000000  ...  0.502793  0.092699  0.107233
...         ...   ...       ...  ...       ...       ...       ...
43147  0.066528  0.75  0.666667  ...  0.482309  0.087267  0.098428
43148  0.103950  0.25  1.000000  ...  0.520484  0.093548  0.110377
43149  0.024948  0.75  0.166667  ...  0.405028  0.074533  0.084277
43150  0.043659  1.00  0.833333  ...  0.447858  0.081324  0.092138
43151  0.147609  1.00  0.666667  ...  0.581006  0.105093  0.119497

[43152 rows x 10 columns]

encoding_results = encoder.transform(test_set[['cut', 'color', 'clarity']])

[[4. 3. 1.]
 [4. 3. 2.]
 [3. 1. 2.]
 ...
 [2. 4. 1.]
 [4. 5. 6.]
 [1. 0. 5.]]

test_set.loc[:, ['cut', 'color', 'clarity']] = encoding_results

       carat  cut  color  clarity  depth  table  price     x     y     z
10176   1.10  4.0    3.0      1.0   62.0   55.0   4733  6.61  6.65  4.11
16083   1.29  4.0    3.0      2.0   62.6   56.0   6424  6.96  6.93  4.35
13420   1.20  3.0    1.0      2.0   61.1   58.0   5510  6.88  6.80  4.18
20407   1.50  4.0    4.0      2.0   60.9   56.0   8770  7.43  7.36  4.50
8909    0.90  2.0    4.0      3.0   61.7   57.0   4493  6.17  6.21  3.82
...      ...  ...    ...      ...    ...    ...    ...   ...   ...   ...
42208   0.52  1.0    3.0      3.0   63.6   57.0   1289  5.05  5.10  3.23
3638    0.91  2.0    5.0      1.0   60.4   61.0   3435  6.21  6.28  3.77
5508    1.08  2.0    4.0      1.0   63.4   55.0   3847  6.53  6.50  4.13
19535   1.02  4.0    5.0      6.0   61.5   57.0   8168  6.44  6.47  3.97
47950   0.50  1.0    0.0      5.0   64.8   58.0   1917  4.97  5.00  3.23

[10788 rows x 10 columns]

test_set = normaliser.fit_transform(test_set)

[[0.22900763 1.         0.5        ... 0.661      0.6751269  0.63919129]
 [0.27735369 1.         0.5        ... 0.696      0.7035533  0.67651633]
 [0.25445293 0.75       0.16666667 ... 0.688      0.69035533 0.65007776]
 ...
 [0.22391858 0.5        0.66666667 ... 0.653      0.65989848 0.64230171]
 [0.2086514  1.         0.83333333 ... 0.644      0.65685279 0.61741835]
 [0.07633588 0.25       0.         ... 0.497      0.50761421 0.50233281]]