Data Analysis Over 10 years of hourly energy consumption using Python Seaborn Pandas

Hourly Energy Consumption¶

Step 1:¶

Import Library¶

In [11]:

import pandas as pd
import numpy as np
import seaborn as sns
import os
import datetime
%matplotlib inline

Step 2:¶

Read the DataSet¶

In [6]:

df = pd.read_csv("AEP_hourly.csv")
df.head(3)

Out[6]:

	Datetime	AEP_MW
0	2004-12-31 01:00:00	13478.0
1	2004-12-31 02:00:00	12865.0
2	2004-12-31 03:00:00	12577.0

In [7]:

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 121273 entries, 0 to 121272
Data columns (total 2 columns):
Datetime    121273 non-null object
AEP_MW      121273 non-null float64
dtypes: float64(1), object(1)
memory usage: 1.9+ MB

In [8]:

df.describe()

Out[8]:

	AEP_MW
count	121273.000000
mean	15499.513717
std	2591.399065
min	9581.000000
25%	13630.000000
50%	15310.000000
75%	17200.000000
max	25695.000000

Step 3:¶

seperate date and time¶

In [18]:

df["New_Date"] = pd.to_datetime(df["Datetime"]).dt.date
df["New_Time"] = pd.to_datetime(df["Datetime"]).dt.time

In [20]:

df1 = df
df1.head(2)

Out[20]:

	Datetime	AEP_MW	New_Date	New_Time
0	2004-12-31 01:00:00	13478.0	2004-12-31	01:00:00
1	2004-12-31 02:00:00	12865.0	2004-12-31	02:00:00

Step 4:¶

When was the higest Energy Consumption and which Year¶

Maximum¶

In [21]:

df1[df1["AEP_MW"] == df["AEP_MW"].max()]

Out[21]:

	Datetime	AEP_MW	New_Date	New_Time
30221	2008-10-20 14:00:00	25695.0	2008-10-20	14:00:00

Minimum¶

In [22]:

df1[df1["AEP_MW"] == df["AEP_MW"].min()]

Out[22]:

	Datetime	AEP_MW	New_Date	New_Time
100759	2016-10-02 05:00:00	9581.0	2016-10-02	05:00:00

Conclusion : From Step 4 we can say that Maximum Energy was Consumed during 2016-10-02 at 05:00:00 and it was 9581.0 MW and Minimum was on 2008-10-20 at 14:00:00 and was 25695.0 MW¶

Step 5:¶

Plot and Data visualization¶

In [24]:

sns.distplot(df1["AEP_MW"])

Out[24]:

<matplotlib.axes._subplots.AxesSubplot at 0x1c2c00b550>

Step 7: Extract Date and Time¶

In [36]:

df1.head(2)

Out[36]:

	Datetime	AEP_MW	New_Date	New_Time
0	2004-12-31 01:00:00	13478.0	2004-12-31	01:00:00
1	2004-12-31 02:00:00	12865.0	2004-12-31	02:00:00

In [38]:

df1["Year"] = pd.DatetimeIndex(df['New_Date']).year

In [39]:

df1.head(2)

Out[39]:

	Datetime	AEP_MW	New_Date	New_Time	Year
0	2004-12-31 01:00:00	13478.0	2004-12-31	01:00:00	2004
1	2004-12-31 02:00:00	12865.0	2004-12-31	02:00:00	2004

Check how many Years are Unique¶

In [41]:

df1["Year"].unique()

Out[41]:

array([2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014,
       2015, 2016, 2017, 2018])

This Tell us that there are 10 Unique Year from 2004 to 2018¶

In [43]:

df1[df1["Year"] == 2013].nunique()

Out[43]:

Datetime    8758
AEP_MW      5601
New_Date     365
New_Time      24
Year           1
dtype: int64

shows the Relationship of Energy vs Year¶

In [44]:

sns.lineplot(x=df1["Year"],y=df1["AEP_MW"], data=df1)

Out[44]:

<matplotlib.axes._subplots.AxesSubplot at 0x1c32358630>

Regression¶

In [48]:

sns.jointplot(x=df1["Year"],
              y=df1["AEP_MW"],
              data=df1,
             kind="reg")

Out[48]:

<seaborn.axisgrid.JointGrid at 0x1c2e91dbe0>

In [49]:

sns.jointplot(x=df1["Year"],
              y=df1["AEP_MW"],
              data=df1,
             kind="kde")

Out[49]:

<seaborn.axisgrid.JointGrid at 0x1c2e3e8358>

Let us see the relation between Energy and Time¶

In [51]:

sns.lineplot(x=df1["New_Time"],y=df1["AEP_MW"], data=df1)

Out[51]:

<matplotlib.axes._subplots.AxesSubplot at 0x1c31c3b748>