import pandas as pd
import matplotlib.pyplot as plt

f = "../data/002_processed/esolmet_2018.parquet"
tmx = pd.read_parquet(f)
tmx.info()

<class 'pandas.core.frame.DataFrame'>
DatetimeIndex: 51173 entries, 2018-01-01 00:00:00 to 2018-12-31 23:50:00
Data columns (total 8 columns):
 #   Column  Non-Null Count  Dtype  
---  ------  --------------  -----  
 0   Ib      51173 non-null  float64
 1   Ig      51173 non-null  float64
 2   Id      51173 non-null  float64
 3   uv      51173 non-null  float64
 4   To      51173 non-null  float64
 5   hr      51173 non-null  float64
 6   ws      51173 non-null  float64
 7   p       51173 non-null  float64
dtypes: float64(8)
memory usage: 3.5 MB

tmx

	Ib	Ig	Id	uv	To	hr	ws	p
Fecha
2018-01-01 00:00:00	0.057	0.0	0.0	0.001	18.93	41.57	1.253	879.0692
2018-01-01 00:10:00	0.002	0.0	0.0	0.001	18.76	41.00	0.418	879.4363
2018-01-01 00:20:00	0.170	0.0	0.0	0.001	18.92	40.96	0.955	879.5181
2018-01-01 00:30:00	0.371	0.0	0.0	0.001	18.52	42.46	1.823	879.5826
2018-01-01 00:40:00	0.305	0.0	0.0	0.001	18.49	42.43	2.149	879.6826
...	...	...	...	...	...	...	...	...
2018-12-31 23:10:00	0.125	0.0	0.0	0.000	18.88	59.60	2.145	875.5595
2018-12-31 23:20:00	0.000	0.0	0.0	0.000	18.71	59.67	1.638	875.5595
2018-12-31 23:30:00	0.044	0.0	0.0	0.000	18.52	58.75	1.923	875.2889
2018-12-31 23:40:00	0.170	0.0	0.0	0.000	18.36	60.62	2.089	875.0606
2018-12-31 23:50:00	0.003	0.0	0.0	0.000	17.99	60.76	0.744	875.1424

51173 rows × 8 columns

def clasifica_temperatura(To):
    if To >= 27:
        return 3
    elif To>=25:
        return 2
    elif To>=22:
        return 1
    elif To>=20:
        return 0
    elif To>=18:
        return -1
    elif To>=15:
        return -2
    else:
        return -3
        

clasifica_temperatura(13.5)

-3

tmx["alto_bajo"] = tmx.To.apply(lambda Temp: "3" if Temp>=27 else "2")

tmx["alto_bajo"] = tmx.To.apply(clasifica_temperatura)

resultado = tmx.alto_bajo.value_counts()

resultado

alto_bajo
 3    11714
 1     9822
-1     8333
 0     6916
 2     6701
-2     5638
-3     2049
Name: count, dtype: int64

resultado.sort_index(inplace=True)

resultado

alto_bajo
-3     2049
-2     5638
-1     8333
 0     6916
 1     9822
 2     6701
 3    11714
Name: count, dtype: int64

fig, ax = plt.subplots(figsize=(6,3))

ax.barh(resultado.index,resultado)

ax.grid(alpha=0.2)
ax.set_xlabel("Ocurrencia")
ax.set_ylabel("Nivel calor ")
ax.set_title("Temixco, 2018")

Text(0.5, 1.0, 'Temixco, 2018')