16  Localizar y reemplazar información en Pandas

En esta sesión veremos el uso de la función .loc[ ]. Aprenderás a localizar y modificar información dentro de un DataFrame de manera efectiva. Descubrirás cómo seleccionar datos concretos, segmentar secciones específicas y aplicar filtros para ajustarse a tus requerimientos de análisis de datos.

Utilizala para extraer subconjuntos de filas, eligiendo rangos específicos basados en etiquetas. Establece condiciones para acceder a segmentos del DataFrame que satisfagan criterios definidos, asegurando así la obtención de datos precisos y pertinentes. Prepara tus datos para análisis futuros con .loc[ ], una herramienta esencial para seleccionar, filtrar y modificar la información clave de tu DataFrame.

import pandas as pd

f = "../data/Cuernavaca_Enero_comas.csv"
cuerna = pd.read_csv(f,index_col=0,parse_dates=True)
cuerna.head()

	To	RH	P	Ws	Wd	Ig	Ib	Id
tiempo
2012-01-01 00:00:00	19.3	58	87415	0.0	26	0	0	0
2012-01-01 01:00:00	18.6	59	87602	0.0	26	0	0	0
2012-01-01 02:00:00	17.9	61	87788	0.0	30	0	0	0
2012-01-01 03:00:00	17.3	66	87554	0.0	30	0	0	0
2012-01-01 04:00:00	16.6	71	87321	0.0	27	0	0	0

cuerna.loc["2012-1-2"]

	To	RH	P	Ws	Wd	Ig	Ib	Id
tiempo
2012-01-02 00:00:00	15.1	40	87190	1.0	91	0	0	0
2012-01-02 01:00:00	14.7	44	87280	1.0	78	0	0	0
2012-01-02 02:00:00	14.2	47	87319	1.0	330	0	0	0
2012-01-02 03:00:00	14.2	49	87237	0.8	69	0	0	0
2012-01-02 04:00:00	14.9	51	87179	0.5	67	0	0	0
2012-01-02 05:00:00	16.1	51	87147	0.2	360	0	0	0
2012-01-02 06:00:00	15.0	55	87006	0.0	71	0	0	0
2012-01-02 07:00:00	17.0	52	87095	0.0	66	14	42	11
2012-01-02 08:00:00	19.0	49	87185	0.0	79	148	399	44
2012-01-02 09:00:00	23.0	36	87360	2.1	160	347	679	68
2012-01-02 10:00:00	23.5	35	87381	1.5	171	551	882	78
2012-01-02 11:00:00	24.0	34	87403	1.0	130	699	937	86
2012-01-02 12:00:00	25.5	30	87466	1.0	191	783	959	90
2012-01-02 13:00:00	27.0	26	87530	1.0	150	795	958	93
2012-01-02 14:00:00	27.0	24	87530	1.0	205	681	768	157
2012-01-02 15:00:00	27.0	23	87530	1.0	150	553	712	154
2012-01-02 16:00:00	26.0	26	87488	0.0	212	423	778	82
2012-01-02 17:00:00	25.0	30	87445	0.0	216	213	566	58
2012-01-02 18:00:00	19.4	45	86992	0.4	229	0	0	0
2012-01-02 19:00:00	18.4	47	86893	0.5	239	0	0	0
2012-01-02 20:00:00	17.5	49	86811	0.6	320	0	0	0
2012-01-02 21:00:00	15.9	53	87027	0.6	244	0	0	0
2012-01-02 22:00:00	14.3	56	87121	0.9	247	0	0	0
2012-01-02 23:00:00	12.6	60	87290	1.0	340	0	0	0

cuerna.loc["2012-01-10":"2012-01-17"]

	To	RH	P	Ws	Wd	Ig	Ib	Id
tiempo
2012-01-10 00:00:00	16.4	42	87080	1.0	16	0	0	0
2012-01-10 01:00:00	15.9	46	87224	1.0	10	0	0	0
2012-01-10 02:00:00	15.5	50	87368	1.0	270	0	0	0
2012-01-10 03:00:00	15.4	52	87277	0.8	10	0	0	0
2012-01-10 04:00:00	15.2	54	87187	0.5	13	0	0	0
...	...	...	...	...	...	...	...	...
2012-01-17 19:00:00	20.9	34	87273	0.2	233	0	0	0
2012-01-17 20:00:00	18.8	37	87186	0.2	279	0	0	0
2012-01-17 21:00:00	18.3	38	87162	0.3	314	0	0	0
2012-01-17 22:00:00	17.9	39	87137	0.4	337	0	0	0
2012-01-17 23:00:00	17.4	41	87112	0.5	348	0	0	0

192 rows × 8 columns

#Si queremos seleccionar donde ocurre algo, por ejemplo donde Ig > 0
cuerna.Ig>0.

tiempo
2012-01-01 00:00:00    False
2012-01-01 01:00:00    False
2012-01-01 02:00:00    False
2012-01-01 03:00:00    False
2012-01-01 04:00:00    False
                       ...  
2012-01-31 19:00:00    False
2012-01-31 20:00:00    False
2012-01-31 21:00:00    False
2012-01-31 22:00:00    False
2012-01-31 23:00:00    False
Name: Ig, Length: 744, dtype: bool

cuerna.loc[cuerna.Ig>0]

	To	RH	P	Ws	Wd	Ig	Ib	Id
tiempo
2012-01-01 07:00:00	18.0	70	87140	0.0	34	20	151	11
2012-01-01 08:00:00	19.0	68	87185	0.0	61	164	522	37
2012-01-01 09:00:00	20.0	60	87229	0.0	95	369	812	58
2012-01-01 10:00:00	20.0	64	87229	1.0	108	568	931	68
2012-01-01 11:00:00	20.0	68	87229	2.1	160	717	981	75
...	...	...	...	...	...	...	...	...
2012-01-31 13:00:00	25.0	19	87445	2.6	180	878	990	90
2012-01-31 14:00:00	26.0	20	87488	3.1	240	827	978	88
2012-01-31 15:00:00	26.0	21	87488	3.1	190	705	941	84
2012-01-31 16:00:00	25.0	22	87445	2.6	210	523	872	75
2012-01-31 17:00:00	23.0	27	87360	0.0	234	301	722	59

341 rows × 8 columns

mascara = cuerna.Ig>0
cuerna.loc[mascara]

	To	RH	P	Ws	Wd	Ig	Ib	Id
tiempo
2012-01-01 07:00:00	18.0	70	87140	0.0	34	20	151	11
2012-01-01 08:00:00	19.0	68	87185	0.0	61	164	522	37
2012-01-01 09:00:00	20.0	60	87229	0.0	95	369	812	58
2012-01-01 10:00:00	20.0	64	87229	1.0	108	568	931	68
2012-01-01 11:00:00	20.0	68	87229	2.1	160	717	981	75
...	...	...	...	...	...	...	...	...
2012-01-31 13:00:00	25.0	19	87445	2.6	180	878	990	90
2012-01-31 14:00:00	26.0	20	87488	3.1	240	827	978	88
2012-01-31 15:00:00	26.0	21	87488	3.1	190	705	941	84
2012-01-31 16:00:00	25.0	22	87445	2.6	210	523	872	75
2012-01-31 17:00:00	23.0	27	87360	0.0	234	301	722	59

341 rows × 8 columns

mascara = (cuerna.Ig>0) & (cuerna.Ws != 0)
cuerna.loc[mascara]

	To	RH	P	Ws	Wd	Ig	Ib	Id
tiempo
2012-01-01 10:00:00	20.0	64	87229	1.0	108	568	931	68
2012-01-01 11:00:00	20.0	68	87229	2.1	160	717	981	75
2012-01-01 12:00:00	21.0	60	87273	1.8	135	800	999	79
2012-01-01 13:00:00	22.0	53	87316	1.5	160	810	998	80
2012-01-01 14:00:00	21.7	53	87302	1.3	164	747	977	79
...	...	...	...	...	...	...	...	...
2012-01-31 09:00:00	20.0	30	87229	3.1	150	377	786	62
2012-01-31 13:00:00	25.0	19	87445	2.6	180	878	990	90
2012-01-31 14:00:00	26.0	20	87488	3.1	240	827	978	88
2012-01-31 15:00:00	26.0	21	87488	3.1	190	705	941	84
2012-01-31 16:00:00	25.0	22	87445	2.6	210	523	872	75

181 rows × 8 columns

cuerna.loc[cuerna.Ig==0,"Ig"] = -999

cuerna.Ig.plot()

cuerna.head()

	To	RH	P	Ws	Wd	Ig	Ib	Id
tiempo
2012-01-01 00:00:00	19.3	58	87415	0.0	26	-999	0	0
2012-01-01 01:00:00	18.6	59	87602	0.0	26	-999	0	0
2012-01-01 02:00:00	17.9	61	87788	0.0	30	-999	0	0
2012-01-01 03:00:00	17.3	66	87554	0.0	30	-999	0	0
2012-01-01 04:00:00	16.6	71	87321	0.0	27	-999	0	0

# .loc no sirve para localizar columnas, eso se hace directamente
# cuerna.loc["To"]

cuerna["To"]

tiempo
2012-01-01 00:00:00    19.3
2012-01-01 01:00:00    18.6
2012-01-01 02:00:00    17.9
2012-01-01 03:00:00    17.3
2012-01-01 04:00:00    16.6
                       ... 
2012-01-31 19:00:00    17.9
2012-01-31 20:00:00    16.9
2012-01-31 21:00:00    16.5
2012-01-31 22:00:00    16.3
2012-01-31 23:00:00    16.2
Name: To, Length: 744, dtype: float64