CODEWITHSUNDEEP
pythonpython-2.7pandasscipy
Nyxynyx
Nyxynyx

Reputation: 63619

Unable to integrate over discrete set of points using Python

I have a Pandas Series with the following values. How can we perform integration to find the area under this plot?

Hour
0.00000    1.195617
0.23990    2.408227
0.47980    1.256069
0.71970    2.227347
0.95960    1.397774
1.19949    1.896309
1.43939    1.309016
1.67929    1.827614
1.91919    1.383252
2.15909    1.630766
2.39899    1.360364
2.63889    1.541367
2.87879    1.560319
3.11869    0.743437
3.35859    1.549370
...
20.39141    2.067811
20.63131    1.938257
20.87121    1.944990
21.11111    1.853212
21.35101    1.702590
21.59091    1.746243
21.83081    2.337570
22.07071    3.773000
22.31061    1.532937
22.55051    1.178040
22.79040    1.850222
23.03030    1.092376
23.27020    1.895959
23.51010    0.966083
23.75000    1.950073
Name: Cost, Length: 100, dtype: float64

Tried the integration function from the cookbook but it's throwing an error 

TypeError: Setting <class 'pandas.core.index.Float64Index'> dtype to anything other than float64 or object is not supported

Upvotes: 0

Views: 175

Answers (1)

nitin
nitin

Reputation: 7358

The recipe in the cookbook requires you to use a pd.TimeSeries and not pd.Series. You can convert your index to a pd.Timestamp like so to get a pd.TimeSeries

i = [0.00000,
0.23990,
0.47980,
0.71970,
0.95960,
1.19949,
1.43939,
1.67929,
1.91919,
2.15909,
2.39899,
2.63889,
2.87879,
3.11869,
3.35859] 
c = [1.195617,
2.408227,
1.256069,
2.227347,
1.397774,
1.896309,
1.309016,
1.827614,
1.383252,
1.630766,
1.360364,
1.541367,
1.560319,
0.743437,
1.549370]

import pandas as pd
s = pd.TimeSeries (index = [pd.Timestamp(k) for k in i], data = c)
s.integrate()

This should yield,

Out[29]:
4.3059284999999994e-09

Upvotes: 1

Related Questions