matplotlib的怪异图

Question

I'm trying to plot a demand profile for heating energy for a specific building with Python and matplotlib . But instead of being a single line it looks like this:

Did anyone ever had plotting results like this? Or does anyone have an idea whats going on here?

The corresponding code fragment is:

for b in list_of_buildings:

    print(b.label, b.Q_Heiz_a, b.Q_Heiz_TT, len(b.lp.heating_list))

    heating_datalist=[]
    for d in range(timesteps):
        b.lp.heating_list[d] = b.lp.heating_list[d]*b.Q_Heiz_TT     
        heating_datalist.append((d, b.lp.heating_list[d]))

        xs_heat = [x[0] for x in heating_datalist]
        ys_heat = [x[1] for x in heating_datalist]
        pyplot.plot(xs_heat, ys_heat, lw=0.5)              

pyplot.title(TT)

#get legend entries from list_of_buildings
list_of_entries = []
for b in list_of_buildings:
    list_of_entries.append(b.label)
pyplot.legend(list_of_entries)          
pyplot.xlabel("[min]")
pyplot.ylabel("[kWh]")

Additional info:

• timesteps is a list like [0.00, 0.01, 0.02, ... , 23.59] - the minutes of the day (24*60 values)
• b.lp.heating_list is a list containing some float values
• b.Q_Heiz_TT is a constant

Answer 1

Based on your information, I have created a minimal example that should reproduce your problem (if not, you may have not explained the problem/parameters in sufficient detail). I'd urge you to create such an example yourself next time, as your question is likely to get ignored without it. The example looks like this:

import numpy as np
import matplotlib.pyplot as plt

N = 24*60
Q_Heiz_TT = 0.5
lp_heating_list = np.random.rand(N)
lp_heating_list = lp_heating_list*Q_Heiz_TT

heating_datalist = []

for d in range(N):
    heating_datalist.append((d, lp_heating_list[d]))
    xs_heat = [x[0] for x in heating_datalist]
    ys_heat = [x[1] for x in heating_datalist]
    plt.plot(xs_heat, ys_heat)

plt.show()

What is going in in here? For each d in range(N) (with N = 24*60 , ie each minute of the day), you plot all values up to and including lp_heating_list[d] versus d . This is because heating_datalist , is appended with the current value of d and corresponding value in lp_heating_list . What you get is 24x60=1440 lines that partially overlap one another. Depending on how your backend is handling things, it may be very slow and start to look messy.

A much better approach would be to simply use

plt.plot(range(timesteps), lp_heating_list)
plt.show()

Which plots only one line, instead of 1440 of them.

Answer 2

I suspect there is an indentation problem in your code.

Try this:

heating_datalist=[]
for d in range(timesteps):
    b.lp.heating_list[d] = b.lp.heating_list[d]*b.Q_Heiz_TT     
    heating_datalist.append((d, b.lp.heating_list[d]))

xs_heat = [x[0] for x in heating_datalist]  # <<<<<<<<
ys_heat = [x[1] for x in heating_datalist]  # <<<<<<<<
pyplot.plot(xs_heat, ys_heat, lw=0.5)       # <<<<<<<<

That way you'll plot only one line per building, which is probably what you want.

Besides, you can use zip to generate x values and y values like this:

xs_heat, ys_heat = zip(*heating_datalist)

This works because zip is it's own inverse!