In [13]:

import numpy as np
import pandas as pd

import matplotlib.pyplot as plt
import seaborn as sns

palette=sns.color_palette()
sns.set_theme()

In [52]:

df=pd.read_csv('DUQ_hourly.csv',
               index_col='Datetime',
               parse_dates=True)

In [18]:

df.info()

<class 'pandas.core.frame.DataFrame'>
DatetimeIndex: 119068 entries, 2005-12-31 01:00:00 to 2018-01-02 00:00:00
Data columns (total 1 columns):
 #   Column  Non-Null Count   Dtype  
---  ------  --------------   -----  
 0   DUQ_MW  119068 non-null  float64
dtypes: float64(1)
memory usage: 1.8 MB

In [37]:

fig,ax=plt.subplots(figsize=(12,8))
plt.title('Duquesne Light Co. Estimated Energe Consumption, 2005-2018',fontsize=16)
df.plot(ax=ax)
plt.ylabel('Energy Consumption (MW)')
plt.legend(['Energy (MW)']);

In [38]:

fig,ax=plt.subplots(figsize=(12,8))
plt.title('Duquesne Light Co. Estimated Energe Consumption, 2008-2010',fontsize=16)
df.loc[(df.index<'1-1-2010') & (df.index>='1-1-2008')].plot(ax=ax)
plt.ylabel('Energy Consumption (MW)')
plt.legend(['Energy (MW)']);

In [39]:

import prophet

In [41]:

train=df.loc[df.index<'1-1-2016']
test=df.loc[df.index>='1-1-2016']

In [45]:

_,ax=plt.subplots()
train.plot(ax=ax)
test.plot(ax=ax,c=palette[1])
plt.legend(['train','test'])

Out[45]:

<matplotlib.legend.Legend at 0x78539505afb0>

In [56]:

prophet_train=train.reset_index().rename(columns={'Datetime':'ds','DUQ_MW':'y'})

In [82]:

p=prophet.Prophet()
p.fit(prophet_train)

DEBUG:cmdstanpy:input tempfile: /tmp/tmp3zbqp_kb/zhf5orsk.json
DEBUG:cmdstanpy:input tempfile: /tmp/tmp3zbqp_kb/kyl4x75k.json
DEBUG:cmdstanpy:idx 0
DEBUG:cmdstanpy:running CmdStan, num_threads: None
DEBUG:cmdstanpy:CmdStan args: ['/usr/local/lib/python3.10/dist-packages/prophet/stan_model/prophet_model.bin', 'random', 'seed=77123', 'data', 'file=/tmp/tmp3zbqp_kb/zhf5orsk.json', 'init=/tmp/tmp3zbqp_kb/kyl4x75k.json', 'output', 'file=/tmp/tmp3zbqp_kb/prophet_modelh577ge8_/prophet_model-20240116034527.csv', 'method=optimize', 'algorithm=lbfgs', 'iter=10000']
03:45:27 - cmdstanpy - INFO - Chain [1] start processing
INFO:cmdstanpy:Chain [1] start processing
03:46:40 - cmdstanpy - INFO - Chain [1] done processing
INFO:cmdstanpy:Chain [1] done processing

Out[82]:

<prophet.forecaster.Prophet at 0x785383ccc040>

In [83]:

prophet_test=test.reset_index().rename(columns={'Datetime':'ds','DUQ_MW':'y'})

In [84]:

forecast=p.predict(prophet_test)

In [69]:

_,ax=plt.subplots(figsize=(12,8))
test.plot(ax=ax)
forecast['yhat'].plot(ax=ax)

Out[69]:

<Axes: xlabel='ds'>

In [86]:

_,ax=plt.subplots(figsize=(12,8))
test.plot(ax=ax,color=palette[1])
p.plot(forecast,ax=ax);

In [87]:

p.plot_components(forecast);

In [80]:

forecast

Out[80]:

	ds	trend	yhat_lower	yhat_upper	trend_lower	trend_upper	additive_terms	additive_terms_lower	additive_terms_upper	daily	...	weekly	weekly_lower	weekly_upper	yearly	yearly_lower	yearly_upper	multiplicative_terms	multiplicative_terms_lower	multiplicative_terms_upper	yhat
0	2016-01-01 00:00:00	1587.411121	1289.872403	1709.200353	1587.411121	1587.411121	-79.404115	-79.404115	-79.404115	-102.704821	...	57.072075	57.072075	57.072075	-33.771369	-33.771369	-33.771369	0.0	0.0	0.0	1508.007006
1	2016-01-01 01:00:00	1587.404299	1208.570563	1637.622988	1587.404299	1587.404299	-171.543092	-171.543092	-171.543092	-194.292211	...	56.436864	56.436864	56.436864	-33.687744	-33.687744	-33.687744	0.0	0.0	0.0	1415.861208
2	2016-01-01 02:00:00	1587.397477	1129.266689	1559.934477	1587.397477	1587.397477	-240.691261	-240.691261	-240.691261	-262.793208	...	55.704201	55.704201	55.704201	-33.602253	-33.602253	-33.602253	0.0	0.0	0.0	1346.706216
3	2016-01-01 03:00:00	1587.390655	1105.440848	1524.564908	1587.390655	1587.390655	-280.398922	-280.398922	-280.398922	-301.740829	...	54.856803	54.856803	54.856803	-33.514896	-33.514896	-33.514896	0.0	0.0	0.0	1306.991733
4	2016-01-01 04:00:00	1587.383833	1081.818916	1512.757601	1587.383833	1587.383833	-288.730027	-288.730027	-288.730027	-309.181109	...	53.876755	53.876755	53.876755	-33.425673	-33.425673	-33.425673	0.0	0.0	0.0	1298.653805
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
22675	2018-08-02 20:00:00	1432.713026	538.241389	3412.151695	15.511093	2872.970023	506.742825	506.742825	506.742825	182.370554	...	58.951510	58.951510	58.951510	265.420762	265.420762	265.420762	0.0	0.0	0.0	1939.455852
22676	2018-08-02 21:00:00	1432.706204	439.222654	3367.968389	15.366671	2873.126313	475.537975	475.537975	475.537975	151.859624	...	58.552356	58.552356	58.552356	265.125995	265.125995	265.125995	0.0	0.0	0.0	1908.244179
22677	2018-08-02 22:00:00	1432.699382	472.561048	3309.789409	15.222249	2873.282603	411.428325	411.428325	411.428325	88.482067	...	58.114950	58.114950	58.114950	264.831308	264.831308	264.831308	0.0	0.0	0.0	1844.127708
22678	2018-08-02 23:00:00	1432.692560	317.263981	3260.508485	15.077827	2873.438893	320.174408	320.174408	320.174408	-1.988590	...	57.626287	57.626287	57.626287	264.536711	264.536711	264.536711	0.0	0.0	0.0	1752.866968
22679	2018-08-03 00:00:00	1432.685738	213.141056	3132.068297	14.933405	2873.595182	218.609464	218.609464	218.609464	-102.704821	...	57.072075	57.072075	57.072075	264.242210	264.242210	264.242210	0.0	0.0	0.0	1651.295202

22680 rows × 22 columns

In [ ]: