CODEWITHSUNDEEP
pythonmatplotlibplotsavetitle
qwertz
qwertz

Reputation: 459

Savefig cuts off title

Hey I try to savefig my plot, but it allways cuts off my title. I think it is because of y=1.05 (to set a distance to the title). I can not fix it. Is there a way to save the entire graph?

time=round(t[time_period],0)
most_sensitive=sorted(most_sensitive)
plt.figure(figsize=(10, 5))
plt.suptitle("Scatterplot "+str(name)+" , "+r'$\Delta$'+"Output , Zeit= "+str(time)+" s",fontsize=20,y=1.05)
figure_colour=["bo","ro","go","yo"]
for i in [1,2,3,4]:
    ax=plt.subplot(2,2,i)
    plt.plot(parm_value[:,most_sensitive[i-1]], Outputdiff[:,most_sensitive[i-1]],figure_colour[i-1])
    ax.set_xlabel(name+"["+str(most_sensitive[i-1])+"] in "+str(unit))
    ax.set_ylabel(r'$\Delta$'+"Output")
    lb, ub = ax.get_xlim( )
    ax.set_xticks( np.linspace(lb, ub, 4 ) )
    lb, ub = ax.get_ylim( )
    ax.set_yticks( np.linspace(lb, ub, 8 ) )
    ax.grid(True)


plt.tight_layout()
newpath = r'C:/Users/Tim_s/Desktop/Daten/'+str(name)+'/'+str(time)+'/'+'scatterplot'+'/'
if not os.path.exists(newpath):
    os.makedirs(newpath)
savefig(newpath+str(name)+'.png')

Upvotes: 45

Views: 58705

Answers (3)

MariusSiuram
MariusSiuram

Reputation: 3634

I don't know if my scenario was the same as yours, but I solved my issue by adding the parameter bbox_inches='tight' to the savefig call.

That may be valuable for people that stumble on this question given its title. It would have been for me...

Upvotes: 109

Martin Evans
Martin Evans

Reputation: 46759

It is difficult to know what you are getting, but the following should help to solve it:

Replace your existing suptitle with:

import matplotlib.pyplot as plt
import numpy as np

name = "test"
unit = 'cms'
most_sensitive = [1, 2, 3, 4, 5]
time = 5 #round(t[time_period],0)
most_sensitive=sorted(most_sensitive)
fig = plt.figure(figsize=(10, 5))
figure_colour=["bo","ro","go","yo"]
plt.suptitle("Scatterplot "+str(name)+" , "+r'$\Delta$'+"Output , Zeit= "+str(time)+" s",fontsize=20, y=0.95)

for i in [1, 2, 3, 4]:
    ax = plt.subplot(2, 2, i)
    #plt.plot(parm_value[:,most_sensitive[i-1]], Outputdiff[:,most_sensitive[i-1]],figure_colour[i-1])
    ax.set_xlabel(name+"["+str(most_sensitive[i-1])+"] in "+str(unit))
    ax.set_ylabel(r'$\Delta$'+"Output")
    lb, ub = ax.get_xlim( )
    ax.set_xticks( np.linspace(lb, ub, 4 ) )
    lb, ub = ax.get_ylim( )
    ax.set_yticks( np.linspace(lb, ub, 8 ) )
    ax.grid(True)

plt.tight_layout()
plt.subplots_adjust(top=0.85)     # Add space at top

newpath = r'C:/Users/Tim_s/Desktop/Daten/'+str(name)+'/'+str(time)+'/'+'scatterplot'+'/'
if not os.path.exists(newpath):
    os.makedirs(newpath)

plt.savefig(newpath+str(name)+'.png')

Giving you:

Matplotlib screenshot

Upvotes: 2

tmdavison
tmdavison

Reputation: 69076

You can control the placement of subplots using plt.subplots_adjust. In this case, the relevant option to adjust is the top.

As well as changing that, you will need to make y in suptitle less than 1 (since that works in figure coordinates - anything > 1 will be off the top of the figure). You could even forget about setting y entirely if you set subplots_adjust correctly.

Note that if you still want tight_layout to control the rest of the subplot placement, you will need to have your subplots_adjust line after tight_layout, or whatever you set there will be overwritten.

(Alternatively, you could set left, right and bottom in subplots_adjust, and remove the need for tight_layout).

Here's an example script (taking the relevant parts from your example):

import matplotlib.pyplot as plt

plt.figure(figsize=(10,5))
name='mdot'
time='918.0'

plt.suptitle("Scatterplot "+str(name)+" , "+r'$\Delta$'+"Output , Zeit= "+str(time)+" s",fontsize=20)

for i in [1,2,3,4]:
    ax=plt.subplot(2,2,i)

plt.tight_layout()
plt.subplots_adjust(top=0.88)

plt.savefig('example.png')

enter image description here

Upvotes: 11

Related Questions