Quick intro to matplotlib¶
MCS 275 Spring 2022 - Emily Dumas¶
This is a quick tour of basic plotting with matplotlib. For more detail see:
You can install matplotlib on your own machine (e.g. python3 -m pip install matplotlib).
Matplotlib is often most used in a notebook setting.
Matplotlib and numpy are pre-installed in the Google Colab notebook environment.
Important shift in thinking¶
When working with functions, you are probably used to thinking of x as either a variable or a number.
When working with numpy, most things are vectorized. Taking advantage of that idea, instead of x being just a number, in this notebook we will instead use that name for vector containing all the numbers we plan to consider.
We could call it something else, like xvalues or xvec, but in this notebook we chose the shortest reasonable name, simply x.
Import matplotlib¶
In [4]:
# This submodule is all you need in most cases, and plt is a common
# abbreviated name to use for it.
import matplotlib.pyplot as plt
# For a very few things you need access to the parent module
# This also lets us check the installed version.
# Uncomment the next two lines to import that.
# import matplotlib as mpl
# mpl.__version__
In [5]:
import numpy as np
In [ ]:
#plt.style.available
In [ ]:
#plt.style.use("seaborn-whitegrid")
First line plot¶
In [6]:
# Let's plot y=sin(x)
# make a vector of 100 evenly spaced floats between 0 and 4pi
x = np.linspace(0,4*np.pi,100)
#y = np.array( [ np.sin(t) for t in x] ) # Don't do this, please
y = np.sin(x) # Instead, do this.
plt.figure(figsize=(8,6)) # begin a new figure ( might contain many plots )
plt.plot(x,y) # plot( vector of x vals, vector of y vals )
plt.title("Sine function")
plt.xlabel("x")
plt.ylabel("sin(x)")
plt.show() # show it to me.
In [7]:
# Let's plot y=sin(x)
# make a vector of 100 evenly spaced floats between 0 and 4pi
x = np.linspace(0,4*np.pi,100)
#y = np.array( [ np.sin(t) for t in x] ) # Don't do this, please
y = np.sin(x) # Instead, do this.
plt.figure(figsize=(8,6)) # begin a new figure ( might contain many plots )
plt.plot(x,y,marker="x") # plot( vector of x vals, vector of y vals )
plt.title("Sine function")
plt.xlabel("x")
plt.ylabel("sin(x)")
plt.show() # show it to me.
In [8]:
x = np.linspace(start=-3,stop=3,num=100)
y = np.exp(-x*x)
plt.plot(x,y)
plt.show()
Multiple plots and styling¶
In [13]:
x = np.linspace(start=-3,stop=3,num=500)
y = np.exp(-x*x) # f(x) = e^{-x^2}
y2 = 1.5*np.exp(-3*(x-2)**2) # g(x) = 1.5* e^{-3(x-2)^2}
y3 = 0.8*np.exp(-4*(x+1)**2) # h(x) = 0.8* e^{-4(x+1)^2}
plt.figure(figsize=(8,6))
# Multiple calls to plt.plot all end up on the same set of axes
plt.plot(x,y,color="red",label="Stewart")
plt.plot(x,y2,color="pink",label="Tina")
plt.plot(x,y3,color="#C4E434",label="29.5") # hex RRGGBB
plt.plot(x,0.5+0.7*np.sin(25*x),label="ruiner")
plt.legend()
plt.show()
In [45]:
x = np.linspace(start=-3,stop=3,num=50)
y = np.exp(-x*x)
y2 = 1.5*np.exp(-3*(x-2)**2)
y3 = 0.8*np.exp(-4*(x+1)**2)
plt.plot(x,y,color="#FF0000")
plt.plot(x,y2,linewidth=3,linestyle="dashed")
plt.plot(x,y3,linestyle="",marker="o")
plt.show()
Adjusting axes¶
In [24]:
x = np.linspace(start=-3,stop=3,num=100)
y = np.exp(-x*x)
y2 = 1.5*np.exp(-3*(x-2)**2)
y3 = 0.8*np.exp(-4*(x+1)**2)
x2 = np.linspace(start=-2,stop=4,num=150)
y4 = 0.5*np.exp(-(x2-2)**2)*np.sin(6*x2)
plt.plot(x,y)
plt.plot(x,y2)
plt.plot(x,y3)
plt.plot(x2,y4)
plt.xlim(-1,3)
plt.ylim(0,1.6)
plt.show()
Parametric plot¶
In [51]:
t = np.linspace(0,2*np.pi,300)
x = np.cos(t)
y = np.sin(t)
plt.plot(x,y)
plt.plot(np.sin(2*t),np.cos(t))
plt.axis("equal") # aspect ratio of plot matches aspect ratio of limits
Out[51]:
Line plots can lie¶
In [53]:
x = np.linspace(start=-3,stop=3,num=100)
y = np.tan(x)
plt.plot(x,y)
# TODO: Fix.
Out[53]:
Adding a legend¶
In [ ]:
x = np.linspace(start=-3,stop=3,num=100)
y = np.exp(-x*x)
y2 = 1.5*np.exp(-3*(x-2)**2)
y3 = 0.8*np.exp(-4*(x+1)**2)
plt.figure(figsize=(8,6))
plt.plot(x,y,label="$e^{-x^2}$")
plt.plot(x,y2,label="$1.5e^{-3(x-2)^2}$")
plt.plot(x,y3,label="$0.8e^{-4(x+1)^2}$")
plt.legend()
plt.show()
plt.savefig("three_gaussians.png",dpi=300)
plt.savefig("three_gaussians.pdf")
In [ ]:
x = np.linspace(start=-3,stop=3,num=100)
y = np.exp(-x*x)
y2 = 1.5*np.exp(-3*(x-2)**2)
y3 = 0.8*np.exp(-4*(x+1)**2)
plt.figure(figsize=(8,6))
plt.plot(x,y,label="$e^{-x^2}$",color="orange",linestyle="dashed",linewidth=5)
plt.plot(x,y2,label="$1.5e^{-3(x-2)^2}$",color="#FF0080")
plt.plot(x,y3,label="$0.8e^{-4(x+1)^2}$",linestyle="dotted")
plt.legend()
plt.show()
In [ ]:
x = np.linspace(start=-3,stop=3,num=100)
y = np.exp(-x*x)
y2 = 1.5*np.exp(-3*(x-2)**2)
y3 = 0.8*np.exp(-4*(x+1)**2)
plt.figure(figsize=(8,6))
plt.plot(x,y,label="$e^{-x^2}$",color="orange",linestyle="dashed",linewidth=5)
plt.plot(x,y2,label="$1.5e^{-3(x-2)^2}$",color="#FF0080",marker="*")
plt.plot(x,y3,label="$0.8e^{-4(x+1)^2}$",linestyle="dotted")
plt.legend()
plt.show()
Scatter plots¶
In [56]:
# Fundamentally, plt.plot shows the same marker symbol
# (same size, shape, color) at each data point
n = np.array([1,1.5,2,2.5,3.5,5])
t = np.array([1.8,2.6,3.5,4.9,8.8,8.2])
plt.plot(n,t,marker="o",linestyle="",color="orange")
plt.show()
In [60]:
n = np.array([1,1.5,2,2.5,3.5,5])
t = np.array([1.8,2.6,3.5,4.9,8.8,8.2])
s = np.array([0.1,0.1,0.1,0.2,0.2,0.5])
c = np.array([1,2,3,5,8,20])
plt.scatter(n,t,s=250*s,c=c,marker="o",cmap="Pastel2")
plt.colorbar()
Out[60]:
In [61]:
n = np.array([1,1.5,2,2.5,3.5,5])
t = np.array([1.8,2.6,3.5,4.9,8.8,8.2])
s = np.array([0.1,0.1,0.1,0.2,0.2,0.5])
c = np.array([1,2,3,5,8,20])
plt.scatter(n,t,s=250*s,c=c,marker="o",cmap="seismic")
plt.colorbar()
Out[61]:
In [63]:
n = np.array([1,1.5,2,2.5,3.5,5])
t = np.array([1.8,2.6,3.5,4.9,8.8,8.2])
s = np.array([0.1,0.1,0.1,0.2,0.2,0.5])
c = np.array(["red","red","red","blue","blue","blue"])
plt.scatter(n,t,s=250*s,c=c,marker="o")
Out[63]:
In [58]:
plt.colormaps()
Out[58]:
Scatter plot real world data¶
CSV with data about meteorites recovered on earth's surface, adapted from NASA dataset:
In [2]:
import numpy as np
import matplotlib.pyplot as plt
import csv
In [4]:
# This cell just grabs the contents of "meteorites.csv"
# and returns it as a dict mapping column names to vectors of column data
import collections
columns = collections.defaultdict(list)
with open("meteorites.csv","r",newline="") as fp:
rdr = csv.DictReader(fp)
for row in rdr:
for k in row:
columns[k].append(row[k])
for k in columns:
if k != "year":
columns[k] = np.array(columns[k]).astype("float64")
else:
columns[k] = np.array(columns[k]).astype("int")
In [5]:
plt.figure(figsize=(15,10))
plt.scatter(
columns["longitude"],
columns["latitude"],
s=0.002*columns["mass"]**(0.66), # s sets area of the dot
alpha=0.6,
c=columns["year"],
cmap="PuOr"
)
plt.colorbar()
plt.show()
Pretty nice. Looks like a world map. Questions:
- How much influence does population density have on this?
- What's the likely distribution of actual meteorite landing points?
- What's that really big one from before 1827 near latitude 75?
In [10]:
plt.figure(figsize=(15,10))
plt.scatter(
columns["longitude"],
columns["latitude"],
s=0.002*columns["mass"]**(0.66), # s sets area of the dot
alpha=0.6,
c=columns["year"],
cmap="PuOr"
)
plt.axis("equal")
plt.xlim(-140,-60)
plt.ylim(10,60)
plt.colorbar()
plt.show()
Quick demo of annotate¶
In [7]:
plt.figure(figsize=(15,10))
plt.scatter(
columns["longitude"],
columns["latitude"],
s=0.002*columns["mass"]**(0.66),
alpha=0.6, # 60% opaque, 40% transparent dots.
c=columns["year"],
cmap="PuOr"
)
plt.colorbar()
plt.annotate("Cape York Meteorite (1818)",
xy=(-64.933,76.13), # Point we're annotating
xycoords='data', # inform matlab these coords are in data units
xytext=(0, 15), # Where the text goes
textcoords='offset points', # inform matlab of units and origin for the coords on prev line
# (units = points, origin = the point being annotated)
horizontalalignment='center',
verticalalignment='bottom',
)
plt.show()
That really big meteorite from 1818¶
https://en.wikipedia.org/wiki/Cape_York_meteorite#/media/File:Ahnighito_AMNH,_34_tons_meteorite.jpg
Contour and density plots¶
In [11]:
x = np.linspace(-3,3,100)
y = np.linspace(-2,2,80)
xx,yy = np.meshgrid(x,y)
In [24]:
# f(x,y) = x**3 - 8x + 3*y**2 + 0.5*y**3
zz = xx**3 - 8*xx + 3*yy**2 + 0.5*yy**3 # 80x100 matrix of values of f on the grid
In [25]:
# f(x,y) = 0.2?
plt.figure(figsize=(8,6))
plt.contour(xx,yy,zz,[0.2])
plt.show()
In [31]:
# Contour plot
plt.figure(figsize=(8,6))
plt.contour(xx,yy,zz)
plt.colorbar()
Out[31]:
In [26]:
# Filled contour plot
plt.figure(figsize=(8,6))
plt.contourf(xx,yy,zz)
plt.colorbar()
Out[26]:
Adding labels to contours¶
plt.clabel adds labels to an existing contour plot. Its argument is the return value of a previous call to plt.contour.
In [27]:
plt.figure(figsize=(8,6))
contours = plt.contour(xx,yy,zz,15,cmap="magma")
plt.title("Contour plot")
plt.clabel(contours) # add inline labels to the contours
plt.colorbar()
Out[27]:
Density plots with plt.imshow¶
In [28]:
plt.figure(figsize=(8,6))
plt.imshow(zz,extent=[np.min(x),np.max(x),np.min(y),np.max(y)],origin="lower")
# origin="lower" means the first row of zz appears at the bottom of the plot.
# That's correct since our meshgrid has smallest y values in the first row.
plt.title("Density plot")
plt.colorbar()
Out[28]:
In [29]:
plt.figure(figsize=(8,6))
contours = plt.contour(xx,yy,zz,15,colors="white")
plt.title("Contour and density plot")
plt.clabel(contours) # add inline labels to the contours
plt.imshow(zz,extent=[np.min(x),np.max(x),np.min(y),np.max(y)],origin="lower")
plt.colorbar()
Out[29]:
Aside: UIC's colors¶
In [30]:
# This is adapted from an example in the matplotlib docs:
# https://matplotlib.org/stable/gallery/color/named_colors.html
# This uses lots of matplotlib features we don't cover in MCS 275!
import matplotlib.pyplot as plt
import matplotlib.colors as mcolors
def plot_colortable(colors, title, sort_colors=True, emptycols=0):
cell_width = 212
cell_height = 48
swatch_width = 64
margin = 12
topmargin = 56
# Sort colors by hue, saturation, value and name.
if sort_colors is True:
by_hsv = sorted((tuple(mcolors.rgb_to_hsv(mcolors.to_rgb(color))),
name)
for name, color in colors.items())
names = [name for hsv, name in by_hsv]
else:
names = list(colors)
n = len(names)
ncols = 4 - emptycols
nrows = n // ncols + int(n % ncols > 0)
width = cell_width * 4 + 2 * margin
height = cell_height * nrows + margin + topmargin
dpi = 72
fig, ax = plt.subplots(figsize=(width / dpi, height / dpi), dpi=dpi)
fig.subplots_adjust(margin/width, margin/height,
(width-margin)/width, (height-topmargin)/height)
ax.set_xlim(0, cell_width * 4)
ax.set_ylim(cell_height * (nrows-0.5), -cell_height/2.)
ax.yaxis.set_visible(False)
ax.xaxis.set_visible(False)
ax.set_axis_off()
ax.set_title(title, fontsize=24, loc="left", pad=10)
for i, name in enumerate(names):
row = i % nrows
col = i // nrows
y = row * cell_height
swatch_start_x = cell_width * col
swatch_end_x = cell_width * col + swatch_width
text_pos_x = cell_width * col + swatch_width + 7
ax.text(text_pos_x, y, name+"\n"+colors[name], fontsize=14,
horizontalalignment='left',
verticalalignment='center')
ax.hlines(y, swatch_start_x, swatch_end_x,
color=colors[name], linewidth=28)
return fig
plot_colortable(
{
"Fire Engine Red":"#D50032",
"Navy Pier Blue":"#001E62"
},
"Primary",
sort_colors=False,
emptycols=1
)
plot_colortable(
{
"Chicago Blue":"#41B6E6",
"Champions Gold":"#FFBF3F",
"UI Health Sky Blue":"#0065AD",
"Expo White":"#F2F7EB",
"Steel Gray":"#333333",
},
"Primary",
sort_colors=False,
emptycols=1
)
plt.show()
