matplotlib.pyplot笔记

matplotlib.pyplot Notebook

由于使用emacs-org进行编辑，为方便暂且使用英文

Table of Contents basic point basic elements line graph axis: line markers method 1 method 2 method 3 legend, label and title grid get attributes histogram return instance pie chart instance bar graph instance distinguish between bar-graph and histogram scatter parameters instance subplot method subplot() method subplots() nrows, ncols: sharex, sharey: squeeze: return instance image fill stuffs colors xkcd-mode basic point

hold: whether to hold all lines on same picture,
or del previous lines when drawing new ones

basic elements

axis: x, y, banners, scales, limits

gridline: .lw: linewidth, .alpha: alpha

line graph axis: # simply use axis() plt.axis([0, 10, 0, 1]) # or use following methods plt.xlim(0, 10) plt.ylim(0, 1)

line markers

linestyles

’-’ solid line style

’–’ dashed line style

’-.’ dash-dot line style

’:’ dotted line style

dot-markers

’.’ point marker

’,’ pixel marker

’o’ circle marker

’v’ triangle_down marker

’^’ triangle_up marker

’<’ triangle_left marker

’>’ triangle_right marker

’1’ tri_down marker

’2’ tri_up marker

’3’ tri_left marker

’4’ tri_right marker

’s’ square marker

’p’ pentagon marker

‘*’ star marker

’h’ hexagon1 marker

’H’ hexagon2 marker

’+’ plus marker

’x’ x marker

’D’ diamond marker

’d’ thin_diamond marker

’|’ vline marker

’_’ hline marker

colors

’b’ blue

’r’ red

’g’ green

’c’ cyan

’m’ magenta

’y’ yellow

’w’ white

’k’ black

linewidth

(real number)

method 1

directly set in plot()

plt.plot(x, y, ',:m', linewidth = 0.2) plt.plot(x, y, color = 'm', linestyle = ':', marker = ',', linewidth = 0.2) plt.plot(x, y, color = 'magenta', linestyle = 'dotted', marker = 'pixel', linewidth = 0.2) method 2

set with set()

line = plt.plot(range(5))[0] line.set(color = 'm', linewidth = 0.2, ) method 3

set multiple lines with setp()

lines = plt.plot(range(5), range(8, 13)) plt.setp(lines, color = 'm', linewidth = 0.2, ) legend, label and title

remember to use legend() to activate labels

plt.plot(x, y, label = 'y = sinx')# set a legend plt.legend()# activate plt.xlabel('x axis') plt.ylabel('y axis') plt.title('sin() graph') grid

plt.grid(True)

get attributes

get_()

or getp()

plt.getp(line) plt.getp(line, 'color')# get referred attribute histogram

method hist()

n, bin, patches = \ hist(x, bins = 10, range = None, normed = False, weights = None, cumulative = False, bottom = None, histtype = 'bar', align = 'mid', orientation = 'vertical', rwidth = None, log = False, color = None, label = None, stacked = False, hold = None, data = None, **kwargs)

bin: how many rectangles

range: which rects to be drawn

normed: make the area sum to 1

weights: array-like, weights of each rect when counting area

bottom: array-like or scalar, move each rect or all rects
up(down) by bottom-value

histtype:

‘bar’: traditional

‘barstacked’: stack up when x is multiple input (list of list)

‘step’: the outline, not filled

‘stepfilled’: outline, filled

align: ‘mid’, ‘left’, ‘right’

orientation: ‘horizontal’, ‘vertical’

rwidth: rectangle width

color: can be a array

stacked: whether to stack up multiple input (x), (overlapped with
‘barstacked’ type)

return

n: list, every bin’s value

bins: number of bins

instance x = np.random.randn(10000) plt.hist([x, x * 1.5, x * 2], bins = 20, histtype = 'barstacked', alpha = 0.7, rwidth = 0.5) plt.show()

pie chart

method pie()

pie(x, explode = None, labels = None, colors = {'b', 'g', 'r', 'c', 'm' ,'y', 'k','w'}, autopct = None, pctdistance = 0.6, shadow = False, labeldistance = 1.1, startangle = None, radius = None, counterclock = True, wedgeprops = None, textprops = None, center = (0, 0), frame = False, hold = None, data = None)

explode: array, split the pie XD

colors: color sequence

labels: string sequence

autopct: format string or format function,
to demonstrate the number of each wedge, formatstring % number

pctdistance: how far is the autopct, relevant dist

labeldistance: relevant dist

frame: whether to show axes

instance x = [1, 2, 3, 4] labels = ['one', 'two', 'three', 'four'] explode = [0.1, 0, 0.2, 0] plt.pie(x, labels = labels, explode = explode, autopct = 'take %.1d part', shadow = True) plt.show()

bar graph

method bar()

instance x1 = [1, 3, 5] y1 = [11, 20, 19] x2 = [2, 4, 6] y2 = [10, 5, 6] plt.bar(x1, y1, color = 'm') plt.bar(x2, y2, align = 'center') plt.show()

distinguish between bar-graph and histogram

bar-graph is almost the same to the simple line graph,
though discrete.

histogram is for frequency statistics.

scatter

method scatter

parameters

s: scalar or array --each point’s size

c: color, sequence, or color sequence

instance x = np.random.rand(50) y = np.random.rand(50) x2 = np.arange(-1, 1, 0.1) y2 = np.random.rand(20) s = (20 * np.random.rand(50)) ** 2 c = s s2 = (20 * np.random.rand(20)) ** 2 c2 = s2 plt.scatter(x, y, s = s, c = c, alpha = 0.6) plt.scatter(x2, y2, s = s2, c = c2, alpha = 0.6, marker = '^') plt.show()

subplot method subplot()

three parameters

instance1 below --basic operation

p1 = plt.subplot(2, 1, 1, facecolor = 'k') # when less than 10, can use subplot(211) p2 = plt.subplot(223, facecolor = 'b') p3 = plt.subplot(224) # when counting smaller plots, imagine to split previous bigger ones p1.plot(x, y1) p2.plot(x, y2, 'w') p3.plot(x, y3) plt.show()

instance2 below –twinx(), twiny()

p1 = plt.subplot(111) p2 = p1.twinx() p1.plot(x, y1, 'r') p2.plot(x, y2, 'r') plt.show()

method subplots() f, (p1, p2, ) = \ plt.subplots(nrows = 1, ncols = 1, sharex = False, sharey = False, squeeze = True, subplot_kw = None, **fig_kw) nrows, ncols:

the row/col number for the subplot matrix

sharex, sharey:

True, False

or ‘all’ or 'none’
(same to True or False)

or ‘row’ or 'col’
(when all subplot on same row/col share x/y axis)

squeeze:

bool, when returning, whether to squeeze the dimension

return

figure and a tuple containing each plot

instance f, (p1, p2) = plt.subplots(1, 2, sharey = 'row') p1.plot(x, y1) p2.scatter(x, y2) plt.show() image x = np.random.randn(81).reshape([9, 9]) f, p = plt.subplots(ncols = 3, sharey = True, figsize = [12, 4]) p[0].imshow(x, interpolation = 'nearest') p[0].set_title('nearest') p[1].imshow(x, interpolation = 'bicubic') p[1].set_title('bicubic') p[2].imshow(x, interpolation = 'bilinear') p[2].set_title('bilinear') plt.show()

fill

snowflake instance

import numpy as np import matplotlib.pyplot as plt n = 0 x_list = [] y_list = [] mat = np.array([[1 / 2, -1 / (2 * 3 ** 0.5)], [1 / (2 * 3 ** 0.5), 1 / 2]]) # turn a 30-degree angle def koch_snowflake(cur, p1, p2): global n if cur > n: return p1 = np.array(p1) p2 = np.array(p2) x_list.append(p1[0]) y_list.append(p1[1]) p3 = p1 + (p2 - p1) / 3 koch_snowflake(cur + 1, p1, p3) x_list.append(p3[0]) y_list.append(p3[1]) p4 = p1 + np.matmul(mat, (p2 - p1)) koch_snowflake(cur + 1, p3, p4) x_list.append(p4[0]) y_list.append(p4[1]) p5 = p2 - (p2 - p1) / 3 koch_snowflake(cur + 1, p4, p5) x_list.append(p5[0]) y_list.append(p5[1]) koch_snowflake(cur + 1, p5, p2) x_list.append(p2[0]) y_list.append(p2[1]) def main(): global n n = int(input('please input the recursion depth: ')) koch_snowflake(1, (0, 0), (10, 10 * 3 ** 0.5)) koch_snowflake(1, (10, 10 * 3 ** 0.5), (20, 0)) koch_snowflake(1, (20, 0), (0, 0)) p1 = plt.subplot(111, facecolor = 'teal') plt.axis('equal') p1.fill(x_list, y_list, color = 'w') plt.show() if __name__ == '__main__': main()

stuffs plt.axhspan(10, 20, facecolor = 'b', alpha = 0.3) plt.axvspan(10, 20, facecolor = 'r', alpha = 0.3) plt.axhline(y = 12, color = 'g', xmax = 20) plt.axvline(x = 14, ymin = 0) plt.show()

colors

xkcd-mode

everything old, but with plt.xkcd():

x = np.arange(0, pi * 2, 0.005) y = [sin(xx) for xx in x] with plt.xkcd(): plt.plot(x, y, 'm') plt.title('sin graph with xkcd') plt.show()

作者：LanderXX

pyplot matplotlib