Feedback

×

• Feedback
• Was this page helpful?*
  • Yes
  • No
• Do you have any additional feedback?
  • Yes
  • No
• Additional Feedback

verticapy.machine_learning.model_selection.randomized_features_search_cv

verticapy.machine_learning.model_selection.randomized_features_search_cv(estimator: VerticaModel, input_relation: Annotated[str | vDataFrame, ''], X: Annotated[str | list[str], 'STRING representing one column or a list of columns'], y: str, metric: str = 'auto', cv: int = 3, average: Literal['binary', 'micro', 'macro', 'weighted'] = 'weighted', pos_label: Annotated[bool | float | str | timedelta | datetime, 'Python Scalar'] | None = None, cutoff: Annotated[int | float | Decimal, 'Python Numbers'] = -1, training_score: bool = True, comb_limit: int = 100, skip_error: bool = True, print_info: bool = True, **kwargs) → TableSample

Computes the k-fold grid search of an estimator using different feature combinations. It can be used to find the set of variables that will optimize the model.

estimator: VerticaModel

Vertica estimator with a fit method.

input_relation: SQLRelation

Relation used to train the model.

X: SQLColumns

List of the predictor columns.

y: str

Response Column.

metric: str, optional

Metric used for the model evaluation.

• auto:

  logloss for classification & rmse for regression.

For Classification:

• accuracy:

  Accuracy.

  \[Accuracy = \frac{TP + TN}{TP + TN + FP + FN}\]

• auc:

  Area Under the Curve (ROC).

  \[AUC = \int_{0}^{1} TPR(FPR) \, dFPR\]

• ba:

  Balanced Accuracy.

  \[BA = \frac{TPR + TNR}{2}\]

• bm:

  Informedness

  \[BM = TPR + TNR - 1\]

• csi:

  Critical Success Index

  \[index = \frac{TP}{TP + FN + FP}\]

• f1:

  F1 Score .. math:

  F_1 Score = 2 \times 
  

rac{Precision times Recall}{Precision + Recall}

• fdr:

  False Discovery Rate

  \[FDR = 1 - PPV\]

• fm:

  Fowlkes-Mallows index

  \[FM = \sqrt{PPV * TPR}\]

• fnr:

  False Negative Rate

  \[FNR = \frac{FN}{FN + TP}\]

• for:

  False Omission Rate

  \[FOR = 1 - NPV\]

• fpr:

  False Positive Rate

  \[FPR = \frac{FP}{FP + TN}\]

• logloss:

  Log Loss

  \[Loss = -\frac{1}{N} \sum_{i=1}^{N} \left( y_i \log(p_i) + (1 - y_i) \log(1 - p_i) \right)\]

• lr+:

  Positive Likelihood Ratio.

  \[LR+ = \frac{TPR}{FPR}\]

• lr-:

  Negative Likelihood Ratio.

  \[LR- = \frac{FNR}{TNR}\]

• dor:

  Diagnostic Odds Ratio.

  \[DOR = \frac{TP \times TN}{FP \times FN}\]

• mcc:

  Matthews Correlation Coefficient

• mk:

  Markedness

  \[MK = PPV + NPV - 1\]

• npv:

  Negative Predictive Value

  \[NPV = \frac{TN}{TN + FN}\]

• prc_auc:

  Area Under the Curve (PRC)

  \[AUC = \int_{0}^{1} Precision(Recall) \, dRecall\]

• precision:

  Precision

  \[TP / (TP + FP)\]

• pt:

  Prevalence Threshold.

  \[\frac{\sqrt{FPR}}{\sqrt{TPR} + \sqrt{FPR}}\]

• recall:

  Recall.

  \[TP / (TP + FN)\]

• specificity:

  Specificity.

  \[TN / (TN + FP)\]

For Regression:

• max:

  Max Error.

  \[ME = \max_{i=1}^{n} \left| y_i - \hat{y}_i \right|\]

• mae:

  Mean Absolute Error.

  \[MAE = \frac{1}{n} \sum_{i=1}^{n} \left| y_i - \hat{y}_i \right|\]

• median:

  Median Absolute Error.

  \[MedAE = \text{median}_{i=1}^{n} \left| y_i - \hat{y}_i \right|\]

• mse:

  Mean Squared Error.

  \[MSE = \frac{1}{n} \sum_{i=1}^{n} \left( y_i - \hat{y}_i \right)^2\]

• msle:

  Mean Squared Log Error.

  \[MSLE = \frac{1}{n} \sum_{i=1}^{n} (\log(1 + y_i) - \log(1 + \hat{y}_i))^2\]

• r2:

  R squared coefficient.

  \[R^2 = 1 - \frac{\sum_{i=1}^{n} (y_i - \hat{y}_i)^2}{\sum_{i=1}^{n} (y_i - \bar{y})^2}\]

• r2a:

  R2 adjusted

  \[\text{Adjusted } R^2 = 1 - \frac{(1 - R^2)(n - 1)}{n - k - 1}\]

• var:

  Explained Variance.

  \[VAR = 1 - \frac{Var(y - \hat{y})}{Var(y)}\]

• rmse:

  Root-mean-squared error

  \[RMSE = \sqrt{\frac{1}{n} \sum_{i=1}^{n} (y_i - \hat{y}_i)^2}\]

cv: int, optional

Number of folds.

average: str, optional

The method used to compute the final score for multiclass-classification.

• binary:

  considers one of the classes as positive and use the binary confusion matrix to compute the score.

• micro:

  positive and negative values globally.

• macro:

  average of the score of each class.

• weighted:

  weighted average of the score of each class.

pos_label: PythonScalar, optional

The main class to be considered as positive (classification only).

cutoff: float, optional

The model cutoff (classification only).

training_score: bool, optional

If set to True, the training score is computed with the validation score.

comb_limit: int, optional

Maximum number of features combinations used to train the model.

skip_error: bool, optional

If set to True and an error occurs, the error is displayed but not raised.

print_info: bool, optional

If set to True, prints the model information at each step.

TableSample

result of the randomized features search.

Let us use a dataset which has a variety of predictors and one value of interest. The Titanic dataset is a good example.

import verticapy.datasets as vpd

data = vpd.load_titanic()

	123 pclass Integer	123 survived Integer	Abc Varchar(164)	Abc sex Varchar(20)	123 age Numeric(8)	123 sibsp Integer	123 parch Integer	Abc ticket Varchar(36)	123 fare Numeric(12)	Abc cabin Varchar(30)	Abc embarked Varchar(20)	Abc boat Varchar(100)	123 body Integer	Abc Varchar(100)
1	1	0		male	71.0	0	0	PC 17609	49.5042	[null]	C	[null]	22
2	1	0		male	45.0	0	0	113784	35.5	T	S	[null]	[null]
3	1	0		male	[null]	0	0	113798	31.0	[null]	S	[null]	[null]
4	1	0		male	17.0	0	0	113059	47.1	[null]	S	[null]	[null]
5	1	0		male	27.0	1	0	13508	136.7792	C89	C	[null]	[null]
6	1	0		male	37.0	1	1	PC 17756	83.1583	E52	C	[null]	[null]
7	1	0		male	31.0	1	0	F.C. 12750	52.0	B71	S	[null]	[null]
8	1	0		male	50.0	1	0	PC 17761	106.425	C86	C	[null]	62
9	1	0		female	36.0	0	0	PC 17531	31.6792	A29	C	[null]	[null]
10	1	0		male	37.0	1	0	113803	53.1	C123	S	[null]	[null]
11	1	0		male	24.0	0	0	PC 17593	79.2	B86	C	[null]	[null]
12	1	0		male	45.0	1	0	36973	83.475	C83	S	[null]	[null]
13	1	0		male	40.0	0	0	112059	0.0	B94	S	[null]	110
14	1	0		male	42.0	0	0	113038	42.5	B11	S	[null]	[null]
15	1	0		male	[null]	0	0	17463	51.8625	E46	S	[null]	[null]
16	1	0		male	42.0	1	0	113789	52.0	[null]	S	[null]	38
17	1	0		male	[null]	0	0	PC 17600	30.6958	[null]	C	14	[null]
18	1	0		male	29.0	0	0	113501	30.0	D6	S	[null]	126
19	1	0		male	46.0	0	0	13050	75.2417	C6	C	[null]	292
20	1	0		male	54.0	0	0	17463	51.8625	E46	S	[null]	175
21	1	0		male	47.0	0	0	113796	42.4	[null]	S	[null]	[null]
22	1	0		male	58.0	0	2	35273	113.275	D48	C	[null]	122
23	1	0		male	45.5	0	0	113043	28.5	C124	S	[null]	166
24	1	0		male	29.0	1	0	113776	66.6	C2	S	[null]	[null]
25	1	0		male	47.0	0	0	110465	52.0	C110	S	[null]	207
26	1	0		male	38.0	0	0	19972	0.0	[null]	S	[null]	[null]
27	1	0		male	22.0	0	0	PC 17760	135.6333	[null]	C	[null]	232
28	1	0		male	31.0	0	0	PC 17590	50.4958	A24	S	[null]	[null]
29	1	0		male	50.0	1	0	13507	55.9	E44	S	[null]	[null]
30	1	0		male	56.0	0	0	17764	30.6958	A7	C	[null]	[null]
31	1	0		male	57.0	1	0	PC 17569	146.5208	B78	C	[null]	[null]
32	1	0		female	63.0	1	0	PC 17483	221.7792	C55 C57	S	[null]	[null]
33	1	0		male	61.0	0	0	36963	32.3208	D50	S	[null]	46
34	1	0		male	21.0	0	1	35281	77.2875	D26	S	[null]	169
35	1	0		male	51.0	0	1	PC 17597	61.3792	[null]	C	[null]	[null]
36	1	1		female	63.0	1	0	13502	77.9583	D7	S	10	[null]
37	1	1		female	32.0	0	0	11813	76.2917	D15	C	8	[null]
38	1	1		female	58.0	0	0	113783	26.55	C103	S	8	[null]
39	1	1		female	44.0	0	0	PC 17610	27.7208	B4	C	6	[null]
40	1	1		female	41.0	0	0	16966	134.5	E40	C	3	[null]
41	1	1		female	53.0	0	0	PC 17606	27.4458	[null]	C	6	[null]
42	1	1		male	36.0	0	1	PC 17755	512.3292	B51 B53 B55	C	3	[null]
43	1	1		female	58.0	0	1	PC 17755	512.3292	B51 B53 B55	C	3	[null]
44	1	1		male	11.0	1	2	113760	120.0	B96 B98	S	4	[null]
45	1	1		female	76.0	1	0	19877	78.85	C46	S	6	[null]
46	1	1		female	[null]	0	1	113505	55.0	E33	S	6	[null]
47	1	1		female	39.0	1	1	PC 17756	83.1583	E49	C	14	[null]
48	1	1		female	27.0	1	2	F.C. 12750	52.0	B71	S	3	[null]
49	1	1		female	[null]	0	0	17421	110.8833	[null]	C	4	[null]
50	1	1		female	35.0	0	0	113503	211.5	C130	C	4	[null]
51	1	1		female	22.0	0	1	112378	59.4	[null]	C	7	[null]
52	1	1		female	25.0	1	0	11765	55.4417	E50	C	5	[null]
53	1	1		male	48.0	1	0	PC 17572	76.7292	D33	C	3	[null]
54	1	1		female	35.0	1	0	36973	83.475	C83	S	D	[null]
55	1	1		male	27.0	0	0	PC 17572	76.7292	D49	C	3	[null]
56	1	1		female	24.0	0	0	11767	83.1583	C54	C	7	[null]
57	1	1		female	52.0	1	1	12749	93.5	B69	S	3	[null]
58	1	1		female	44.0	0	1	111361	57.9792	B18	C	4	[null]
59	1	1		female	15.0	0	1	24160	211.3375	B5	S	2	[null]
60	1	1		male	30.0	1	0	13236	57.75	C78	C	11	[null]
61	1	1		female	31.0	1	0	35273	113.275	D36	C	6	[null]
62	1	1		female	39.0	0	0	PC 17758	108.9	C105	C	8	[null]
63	1	1		female	22.0	0	1	113509	61.9792	B36	C	5	[null]
64	1	1		male	52.0	0	0	113786	30.5	C104	S	6	[null]
65	1	1		female	43.0	0	1	24160	211.3375	B3	S	2	[null]
66	1	1		female	33.0	0	0	110152	86.5	B77	S	8	[null]
67	1	1		male	45.0	1	1	16966	134.5	E34	C	3	[null]
68	1	1		female	40.0	1	1	16966	134.5	E34	C	3	[null]
69	1	1		male	48.0	1	0	19996	52.0	C126	S	5 7	[null]
70	1	1		female	[null]	0	0	PC 17585	79.2	[null]	C	D	[null]
71	1	1		female	35.0	0	0	PC 17755	512.3292	[null]	C	3	[null]
72	1	1		female	60.0	1	0	110813	75.25	D37	C	5	[null]
73	1	1		male	21.0	0	1	PC 17597	61.3792	[null]	C	A	[null]
74	2	0		male	23.0	0	0	C.A. 31030	10.5	[null]	S	[null]	[null]
75	2	0		male	28.0	0	0	244358	26.0	[null]	S	[null]	[null]
76	2	0		male	60.0	1	1	29750	39.0	[null]	S	[null]	[null]
77	2	0		female	44.0	1	0	244252	26.0	[null]	S	[null]	[null]
78	2	0		male	29.0	1	0	2003	26.0	[null]	S	[null]	[null]
79	2	0		male	18.0	0	0	S.O.C. 14879	73.5	[null]	S	[null]	[null]
80	2	0		male	18.0	0	0	S.O.C. 14879	73.5	[null]	S	[null]	[null]
81	2	0		male	54.0	0	0	28403	26.0	[null]	S	[null]	[null]
82	2	0		male	18.0	0	0	236171	13.0	[null]	S	[null]	[null]
83	2	0		male	36.0	0	0	229236	13.0	[null]	S	[null]	236
84	2	0		male	34.0	1	0	28664	21.0	[null]	S	[null]	[null]
85	2	0		male	21.0	1	0	28133	11.5	[null]	S	[null]	[null]
86	2	0		male	21.0	1	0	28134	11.5	[null]	S	[null]	[null]
87	2	0		male	24.0	0	0	233866	13.0	[null]	S	[null]	155
88	2	0		male	34.0	0	0	12233	13.0	[null]	S	[null]	[null]
89	2	0		male	30.0	0	0	250653	13.0	[null]	S	[null]	75
90	2	0		male	44.0	0	0	248746	13.0	[null]	S	[null]	35
91	2	0		male	49.0	1	2	220845	65.0	[null]	S	[null]	[null]
92	2	0		male	21.0	2	0	S.O.C. 14879	73.5	[null]	S	[null]	[null]
93	2	0		male	21.0	0	0	S.O.C. 14879	73.5	[null]	S	[null]	[null]
94	2	0		female	60.0	1	0	24065	26.0	[null]	S	[null]	[null]
95	2	0		male	24.0	2	0	C.A. 31029	31.5	[null]	S	[null]	[null]
96	2	0		male	22.0	2	0	C.A. 31029	31.5	[null]	S	[null]	[null]
97	2	0		male	35.0	0	0	233734	12.35	[null]	Q	[null]	[null]
98	2	0		male	31.0	0	0	C.A. 18723	10.5	[null]	S	[null]	165
99	2	0		male	36.0	0	0	SC/Paris 2163	12.875	D	C	[null]	[null]
100	2	0		male	[null]	0	0	SC/A.3 2861	15.5792	[null]	C	[null]	[null]

Rows: 1-100 | Columns: 14

Note

VerticaPy offers a wide range of sample datasets that are ideal for training and testing purposes. You can explore the full list of available datasets in the Datasets, which provides detailed information on each dataset and how to use them effectively. These datasets are invaluable resources for honing your data analysis and machine learning skills within the VerticaPy environment.

Next, we can initialize a Logistic Regression model:

from verticapy.machine_learning.vertica import LogisticRegression

model = LogisticRegression()

Now we can conveniently use the randomized_features_search_cv function to do either forward or backward randomized features search feature selection.

from verticapy.machine_learning.model_selection import randomized_features_search_cv

result = randomized_features_search_cv(
    model,
    input_relation = data,
    X = ["age", "fare", "parch", "pclass",],
    y = "survived",
    cv = 3,
)

		avg_score	avg_train_score	avg_time	score_std	score_train_std
1		0.2506001261650903	0.26292938633935964	2.214517513910929	0.0021474920385460504	0.0014016221399229076
2		0.25074602292517834	0.258169167403599	2.266688823699951	0.002046988889454484	0.000768794229565933
3		0.25399281520670264	0.25525992478844167	2.2767723401387534	0.00315083091224646	0.0011982141807638281
4		0.25557624306861665	0.25740215279109235	2.407245635986328	0.006920130599150315	0.0031673112602994785
5		0.25653137656482833	0.2544579321287073	2.2309447129567466	0.010381790871957902	0.004152680787457593
6		0.26065839032443366	0.25701399068860564	2.306551138559977	0.00840394245595519	0.004271851904456816
7		0.26068955188810367	0.260449149333355	2.108332395553589	0.005061111436909515	0.00253459894411372
8		0.2677091402658934	0.26917598855734864	2.213731129964193	0.0036364307659896237	0.0015706041108292563
9		0.269796020781651	0.2536089817699007	2.2880465984344482	0.0036296829082345544	0.0016344413855915673
10		0.27025884634895503	0.267955563932531	2.284230391184489	0.006248515359034414	0.002939782840266127
11		0.27064012464444365	0.27283658450289433	2.216388702392578	0.004816805834553563	0.0025788358238585207
12		0.27538037572278934	0.27069129274848	2.378204584121704	0.008289043847921916	0.004532597904482789
13		0.2821382243131407	0.28421288580200804	2.0592734813690186	0.003947380427782839	0.0019378427439336605
14		0.29081522875963367	0.2868269658248837	2.1436642011006675	0.007076125438859293	0.0036100428034907157
15		0.293041322959174	0.28966266215839864	2.1305018266042075	0.0068726562199244	0.002929513909999179

Rows: 1-15 | Columns: 6

Note

The models are arranged in ascending order of avg_score.