Trouble with generating graph from function in python

Question

I want to generate a line graph of temperature versus power and have written the following code:

import numpy as np 
import matplotlib.pyplot as plt

temps = np.arange(288, 314, 1)

sigma = 5.67e-8 #Stefan's constant, units = W m^-2 K^-4
T0 = 298 #ambient T, 25 degrees
emissivity = 0.96
A = (np.pi * 309)/80000 #total surface area of chameleon

def SB(T):
    power = A * sigma * emissivity * (T**4 - T0**4)
    return power

power = SB(temps)

plt.plot(temps, power)
plt.xlabel("Temperature (K)")
plt.ylabel("Power (W $m^{-2}$)")
plt.title("Net Power Radiated vs T")
plt.xlim(288, 314)
plt.ylim(0, 1.2)
plt.show()

The axes are showing but I'm not getting a line, could it be because the function doesn't take values from np.arange? If so how would I fix this? Any help would be appreciated! :D

Rachel

Are you coding in an interactive session? Maybe restart the interpreter. — juanpa.arrivillaga, Mar 01 '17 at 20:34
@juanpa.arrivillaga I'm working in Spyder but have restarted it and tried it in another interpreter as well and it's now saying "RuntimeWarning: invalid value encountered in power" :/ — Rachel Eyles, Mar 01 '17 at 20:46
@juanpa.arrivillaga, yeah Python 2.7 I think, not clued up on the big differences between the two haha — Rachel Eyles, Mar 02 '17 at 09:59

score 0 · Accepted Answer · edited May 23 '17 at 11:46

0

This is a funny error. First thing, you don't see the line, because it's outside the ylim range. Delete the code line which sets the ylim, and you'll see the line.

However the line shows negative values - which is strange, since power should be positive and the formulas are correct.
The reason for that is that the input temperature array temps is an int32 array. This array to the power of four will give negative values, since the numbers are simply too big.

A solution to use a floating point array instead. Either use

temps = np.arange(288., 314., 1)

or

temps = np.arange(288, 314, 1).astype(np.float32)

This is reproduced with python 2.7 and numpy 1.12 on windows 8 64bit. The reason for this strange behaviour is seen in this question: "In Microsoft C, even on a 64 bit system, the size of the long int data type is 32 bits.".

edited May 23 '17 at 11:46

Community

1
1

answered Mar 01 '17 at 20:51

ImportanceOfBeingErnest

321,279
53
665
712

Ah, I think that might be it. On my system, `temp` is dtype('int64'). This depends on whatever `C int` is on your system. – juanpa.arrivillaga Mar 01 '17 at 20:56
Any idea why that would be different? Is it python version, numpy version, operating system? – ImportanceOfBeingErnest Mar 01 '17 at 20:57
It depends on your system architecture (for me, I'm on x86_64) – juanpa.arrivillaga Mar 01 '17 at 20:59
And actually, will depend on your OS, unfortunately. But it is whatever a C `long` is on your sytem. – juanpa.arrivillaga Mar 01 '17 at 21:01
I think [this question](http://stackoverflow.com/questions/36278590/numpy-array-dtype-is-coming-as-int32-by-default-in-a-windows-10-64-bit-machine) provides the answer: *"In Microsoft C, even on a 64 bit system, the size of the long int data type is 32 bits."*. – ImportanceOfBeingErnest Mar 01 '17 at 21:05
Ahh thank you everyone, using a floating point array has solved it :D – Rachel Eyles Mar 02 '17 at 10:21
@Rachel If this solves your issue, consider [accepting] the answer, such that it will not stay in the unsolved list forever. – ImportanceOfBeingErnest Mar 02 '17 at 10:32
@ImportanceOfBeingErnest thanks just done it, site newbie sorry! – Rachel Eyles Mar 03 '17 at 12:30

Trouble with generating graph from function in python

1 Answers1