FeatureSelection.get_X_original

FeatureSelection.get_X_original()

Restituisce il DataFrame originale contenente solo le feature inizialmente considerate.

Seleziona dal DataFrame originale (attributo db) solo le colonne presenti nell’attributo feat_to_check, ovvero le feature che sono state incluse all’inizio del processo di analisi/selezione.

Returns:

Un DataFrame contenente solo le colonne specificate in feat_to_check.

Return type:

pd.DataFrame

Dati utilizzati per gli esempi:

>>> db_test_filters

	feature_B	feature_C	feature_A1	feature_A2
0	34.835708	Z	-0.270712	-0.812137
1	3.086785	X	0.104848	0.314544
2	42.384427	Y	0.250528	0.751583
3	86.151493	X	-0.925200	-2.775600
4	-1.707669	V	0.567144	1.701431
5	-1.706848	V	-1.040180	-3.120541
6	88.960641	X	-0.153676	-0.461028
7	48.371736	V	0.789852	2.369555
8	-13.473719	Z	-1.226216	-3.678648
9	37.128002	Y	-0.948007	-2.844021
10	-13.170885	X	-0.569654	-1.708962
11	-13.286488	Z	-0.977150	-2.931451
12	22.098114	W	-0.770632	-2.311895
13	-85.664012	W	-0.033711	-0.101134
14	-76.245892	X	-1.032859	-3.098578
15	-18.114376	X	1.142427	3.427282
16	-40.641556	Z	-0.609778	-1.829334
17	25.712367	Y	1.469416	4.408249
18	-35.401204	Y	1.492679	4.478037
19	-60.615185	Y	0.707125	2.121376

Esempio:

>>> from cefeste.selection import FeatureSelection
>>> fs = FeatureSelection(
...    db=df_test_filters,
...    verbose=True # Utile per vedere cosa succede
... )
>>> fs.find_correlated_features()
>>> fs.make_report()

	feat_name	result	drop_reason
0	feature_A2	drop	correlated
1	feature_A1	keep	NaN
2	feature_C	keep	NaN
3	feature_B	keep	NaN

>>> fs.get_X_original()

	feature_B	feature_C	feature_A1	feature_A2
0	34.835708	Z	-0.270712	-0.812137
1	3.086785	X	0.104848	0.314544
2	42.384427	Y	0.250528	0.751583
3	86.151493	X	-0.925200	-2.775600
4	-1.707669	V	0.567144	1.701431
5	-1.706848	V	-1.040180	-3.120541
6	88.960641	X	-0.153676	-0.461028
7	48.371736	V	0.789852	2.369555
8	-13.473719	Z	-1.226216	-3.678648
9	37.128002	Y	-0.948007	-2.844021
10	-13.170885	X	-0.569654	-1.708962
11	-13.286488	Z	-0.977150	-2.931451
12	22.098114	W	-0.770632	-2.311895
13	-85.664012	W	-0.033711	-0.101134
14	-76.245892	X	-1.032859	-3.098578
15	-18.114376	X	1.142427	3.427282
16	-40.641556	Z	-0.609778	-1.829334
17	25.712367	Y	1.469416	4.408249
18	-35.401204	Y	1.492679	4.478037
19	-60.615185	Y	0.707125	2.121376