Here is the second lesson in Financial Computing and Programming
What I'm learning today is matplotlib - the most basic drawing library in python.

import matplotlib.pyplot as plt
import numpy as np

1. We first generate a one-dimensional y (to draw a line chart)

Just draw a picture

np.random.seed(250)
y = np.random.standard_normal(20)
x = range(len(y))
plt.plot(x,y)

This is the same as the result of direct plt.plot(y)

2. Let's try the cumsum function initially

#add canvas here
plt.figure(figsize = [7,4])
#This is used to draw straight lines
plt.plot(y.cumsum() , "C1" , lw =1.2)
#here to draw
plt.plot(y.cumsum() , "ro")

#This is used to add gridlines
plt.grid("True")
#Here is used to add x-axis and y-axis
plt.xlabel("index")
plt.ylabel("value")
#Here is used to add a title
plt.title("A simple plot")

For the cumsum function, axis=1 refers to the operation on the column (but it is not clear whether other axes will mean the same thing?)

You can see that the first column is unchanged

#For the cumsum function, axis=0 refers to operating on rows

You can see that the first line is unchanged.

3. Next we generate a two-dimensional y (to draw a line chart)

y = np.random.standard_normal((20,2))
y

plt.plot(y[:,0] , lw =1.2 , label = "1st")
plt.plot(y[:,1] , lw = 1.2 , label = '2nd')

plt.plot(y,"ro")

plt.grid("True")

plt.legend(loc = 0 )  # It is set to 0 here, which means the best match

plt.xlabel("index")
plt.ylabel("value")

plt.title("This is my second picture")
#The following is the function used to store, you can use the parameter dpi, such as =200 to specify the storage resolution
plt.savefig("this_is_my_second_picture.png",dpi = 200)

Here is a description of the loc parameter

3. Next, we will y[:,0] = y[:,0] * 100 (what should we do if we encounter pictures with too large scale difference?)

plt.plot(y)

#If you encounter such a large-scale thing, you cannot place it in a picture.

We consider implementing the idea of ​​"subgraph"

plt.subplot(nrows, ncols, index, **kwargs)


plt.figure(figsize = (7,4))

plt.subplot(211)#The number here means 2 rows, 1 column, and the serial number is 1


plt.plot(y[:,0] , lw = 1.2 , label = "1st")
plt.plot(y,"ro")

plt.grid("True")

plt.legend(loc = 0 )  # It is set to 0 here, which means the best match

plt.xlabel("index")
plt.ylabel("value")




plt.subplot(212)#The number here means 2 rows, 1 column, and the serial number is 2
plt.plot(y[:,1] , lw = 1.2 , label = "1st")
plt.plot(y,"ro")

plt.grid("True")

plt.legend(loc = 0 )  # It is set to 0 here, which means the best match

plt.xlabel("index_2")
plt.ylabel("value_2")

plt.figure(figsize = (7,4))

plt.subplot(121)#The number here means 1 row, 2 columns, and the serial number is 1


plt.plot(y[:,0] , lw = 1.2 , label = "1st")
plt.plot(y,"ro")

plt.grid("True")

plt.legend(loc = 0 )  # It is set to 0 here, which means the best match

plt.xlabel("index")
plt.ylabel("value")




plt.subplot(122)#The number here means 1 row, 2 columns, and the serial number is 2
plt.plot(y[:,1] , lw = 1.2 , label = "1st")
plt.plot(y,"ro")

plt.grid("True")

plt.legend(loc = 0 )  # It is set to 0 here, which means the best match

plt.xlabel("index_2")
plt.ylabel("value_2")

Let's take a look at the statement used to draw the bar

plt.figure(figsize = (7,4))

4. Draw a double-axis graph in an object-oriented way (more recommended)

fig,  ax  = plt.subplots(nrows  = 1  , ncols = 2)
# nrows means rows
# ncols for columns
ax[0].plot([1,2,3],[4,5,6])
ax[1].scatter([1,2,3],[4,5,6])

ax[0].set_xlabel("index-1")
ax[1].set_xlabel("index-2")

ax[0].set_title("line")
ax[1].set_title("scatter")

For the above scales that are too different, if we want to place them in one picture, we can consider using a dual-axis graph
The main use here is

ax2 = ax1.twinx()

Equivalent to overlaying two pictures on top of each other

fig = plt.figure()
ax1 = fig.add_subplot(111)

ax2 = ax1.twinx()

ax1.plot(y[:,0]  , lw =1.2,color = 'g' , label = "1st")

ax1.plot(y[:,0],"ro")

ax1.grid("True")

ax1.legend(loc = 0 )  # It is set to 0 here, which seems to mean the best match.

ax1.set_xlabel("index")

ax1.set_ylabel("value")


ax2.plot(y[:,1]  , lw =1.2,color = 'b' , label = "2nd")

ax2.plot(y[:,1],"ro")

ax2.grid("True")

ax2.legend(loc = 0 )  # It is set to 0 here, which seems to mean the best match.

ax2.set_xlabel("index")

ax2.set_ylabel("value")

5. Further improvements (legend s overlap each other?)

We found that our legend s overlapped with each other, only one. So let's solve this problem

fig = plt.figure()


#here are axes
ax1 = fig.add_subplot(111)
ax2 = ax1.twinx()

lne1 = ax1.plot(y[:,0] , lw = 1.2 , color = "b" , label = "1st")
ax1.plot(y[:,0] , "ro")
ax1.grid()
ax1.set_xlabel("index")
ax1.set_ylabel("this_is_the_values")

ax1.set_title("this is some ?")

# ax1.legend(loc = 2)

lne2 = ax2.plot(y[:,1] , lw = 1.2 , color = "C6"  , label = "2nd")

ax2.set_ylabel("vlaue_2")
# ax2.legend(loc = 1)


# #add Our line
lns = lne1 + lne2

labs = [l.get_label() for l in lns]
print(lns)
print(labs)
ax2.legend(lns,labs,loc= 9)

6. draw bar

plt.bar(x = np.arange(len(y)) , height = y[:,1])

plt.grid("True")

plt.legend(loc = 0 )  # It is set to 0 here, which seems to mean the best match.

plt.xlabel("index")

plt.ylabel("value")

7. Here is the scatter plot:

data = np.random.standard_normal((20,2))
plt.figure(figsize = (8,4))
plt.scatter(data[:,0] , data[:,1] , marker = 'o' , color = "C6")

plt.xlabel('x')
plt.ylabel("y")
plt.grid()
plt.title("Here is the scatter plot")

We further refine it

The c here represents the "frequency" of this two-dimensional point

c = np.random.randint(0,10,  len(y))
c

plt.figure(figsize = (7,4))

plt.scatter(data[:,0] , data[:,1] , c = c , marker = "o")
plt.colorbar()
plt.xlabel('x')
plt.ylabel("y")
plt.grid()
plt.title("Here is a better looking scatter plot")

8. Histogram

plt.figure(figsize = (7,19))
plt.hist(data,label=["1st" , "2nd"] , bins  = 10)

9. Box diagram

plt.boxplot(data)

10. Now let's draw a graph of a quadratic function!

def f(x):
    return x**2 - x*2
x = np.linspace(0,2)
y=f(x)
fig,ax = plt.subplots()
plt.plot(x,y)

11.3D data mapping

strike = np.linspace(50,150,24)
ttm = np.linspace(0.5,2.5,24)
strike1,ttm1 =  np.meshgrid(strike ,ttm)
#numpy.linspace(start, stop, num=50, endpoint=True, retstep=False, dtype=None) returns evenly spaced numbers within the specified interval.

inv = strike1 * ttm1

from mpl_toolkits.mplot3d import Axes3D

fig  = plt.figure(figsize = (14,7)) # This is used to control the size of the entire image


ax= fig.gca(projection = "3d")

surf = ax.plot_surface(strike1 , ttm1 , inv , rstride = 2 , cstride = 2,cmap = plt.cm.coolwarm , linewidth = 3)

ax.set_xlabel("strike")
ax.set_ylabel("ttm1")
ax.set_zlabel("inv")


fig.colorbar(surf , shrink = 0.5 , aspect =5) # The following parameters here are used to adjust the size of the column next to it



plt.savefig("this.png" ,dpi = 200 )