Matplotlib

畫一張空的圖

plt.plot()
plt.show()

加上標題及格線

plt.title("uniform accelerated motion") #標題
plt.xlabel("Time") #x軸標題
plt.ylabel("Velocity") #y軸標題
plt.grid(True) #格線
plt.plot()
plt.show()

調整刻度

plt.xticks([0,1,2,3,4])
plt.yticks([0,0.8,1.0,1.3,1.7])
plt.show()

• plt.axis([xmin, xmax, ymin, ymax])

plt.axis([0,4,0,2])
plt.show()

Case 1: 只給y值

x會自動被設成 0, 1, …, N

plt.plot([0, 1, 4, 9, 16])
plt.show()

Case 2: 給x值和y值

x = [0, 0.2, 0.4, 0.6, 0.8]
y = [ i**2 for i in x ]
plt.plot(x, y)
plt.show()

修改樣式

x = [0, 0.2, 0.4, 0.6, 0.8]
y = [ i**2 for i in x ]
plt.plot(x, y, 'r:', linewidth=5)
plt.show()

r代表紅色(red)，:代表虛線

用 plt.plot? 看看有哪些樣式和顏色

直的

langs = ('Python', 'C', 'Haskell', 'Rust')
idxs = range(len(langs))
popularity = [9,4,8,7]
 
plt.bar(idxs, popularity, align='center')
plt.xticks(idxs, langs)
plt.ylabel('Popularity')
plt.title('Programming languages')
 
plt.show()

橫的

langs = ('Python', 'C', 'Haskell', 'Rust')
idxs = range(len(langs))
popularity = [9,4,8,7]
 
plt.barh(idxs, popularity, align='center')
plt.yticks(idxs, langs)
plt.xlabel('Popularity')
plt.title('Programming languages')
 
plt.show()

兩張圖疊一起

bar_width = 0.25
langs = ('Python', 'C', 'Haskell', 'Rust')
popularity_tw = [9,4,8,7]
popularity_us = [10,5,1,3]
idxs_tw = range(len(langs))
idxs_us = [ x+bar_width for x in idxs_tw ]
idxs_label = [ x+bar_width/2 for x in idxs_tw ]
plt.bar(idxs_tw, popularity_tw, bar_width, align='center', color='g', label='TW')
plt.bar(idxs_us, popularity_us, bar_width, align='center', color='m', label='US')
plt.xticks(idxs_label, langs)
plt.ylabel('Popularity')
plt.title('Programming languages')
plt.legend() #顯示label 
plt.show()

status = ('AC', 'WA', 'RE', 'TLE', 'CE')
count = [11, 4, 9, 3, 5]
plt.pie(count, labels=status, autopct='%1.1f%%',
    startangle=30)
plt.axis('equal') #讓x軸 y軸等比例
plt.show()

分離某塊

status = ('AC', 'WA', 'RE', 'TLE', 'CE')
explode = (0.1, 0, 0, 0, 0)
count = [11, 4, 9, 3, 5]
plt.pie(count, explode=explode, labels=status,
    autopct='%1.1f%%', startangle=30)
plt.axis('equal') #讓x軸 y軸等比例
plt.show()

import matplotlib.animation as animation

animation.FuncAnimation(fig, func, frames=frames, 
    init_func=init_func, interval=interval)

• func: 每次更新畫面之函數
• frames: iterable (像list的東西)
• init_func: 初始函數
• interval: 每隔多少ms更新一次畫面

流程圖

範例

import matplotlib.animation as animation
import math


def update(i):
    line.set_ydata([ math.sin(v + i*math.pi/10) for v in x ])
    return [line]

def init():
    return [line]

x = []
v = 0
while v < math.pi * 6:
    x.append(v)
    v += 0.01
y = [ math.sin(v) for v in x ]

fig, ax = plt.subplots()
[line] = ax.plot(x, y)
ax.axis([min(x), max(x), -3, 3])

ani = animation.FuncAnimation(fig, update, frames=range(0,200), interval=25, repeat=True, init_func=init)
plt.show()

• 動畫無法直接在jupyter notebook裡面看，請使用terminal
• 或轉成HTML5影片

from IPython.display import HTML
HTML(ani.to_html5_video())

說明

給你一個neoj的題號，計算該題每種結果(AC,WA etc.)的人數有多少，並畫成圓餅圖

API

傳入一個題號，此函數會回傳該題在NEOJ上的submission的list

import requests
def get_neoj_status(probid):
    ret = []
    count = 1
    while len(ret) < count:
        params = {'offset': len(ret), 'filter': {'problem_uid': probid}}
        r = requests.post('https://neoj.sprout.tw/api/challenge/list', json=params)
        result = r.json()
        ret.extend(result['data'])
        count = result['count']
    return ret

其中每一個submission長得如下:

{'metadata': {'memory': 30692, 'result': 3, 'runtime': 120},
 'problem': {'checker': 'diff',
  'lang': 'python3',
  'memlimit': 65536,
  'name': '修羅道的開始',
  'revision': '351559ab22d93f9a01945b4a49f5befce015624f',
  'subtask': [10, 10, 10, 10, 10, 10, 10, 10, 10, 10],
  'timelimit': 1000,
  'uid': 100},
 'state': 2,
 'submitter': {'category': 0, 'name': 'Administrator', 'uid': 1},
 'timestamp': '2017-02-22T02:32:29.784723+00:00',
 'uid': 1}

['metadata']['result']就是該submission的結果

結果對照表:

1. Accepted
2. Wrong Answer
3. Runtime Error
4. Time Limit Exceed
5. Memory Limit Exceed
6. Compile Error
7. Other