"""

读取txt文件，取有效数据，绘制分流排名分布图。

"""

import numpy as np

import matplotlib.pyplot as plt

import seaborn as sns

import pandas as pd

filename = 'data/分流专业排名2020.txt'

fig_title = "工科试验班2019级专业分流推导排名分布图（仅供参考）"

group_distance = 20 # 界限排名属于排名较高的组

max_rank = 256

# 先读取一个原始表，每个数据结构专业-[所有有效排名]

ranks = {}

with open(filename,errors='ignore') as fp:

for line in fp:

ls = line.strip().split()

if len(ls)!=2:

continue

try:

rank=int(ls[1])

except ValueError:

continue

else:

ranks.setdefault(ls[0],[]).append(rank)

for lst in ranks.values():

lst.sort()

groups = {}

lst=[]

for major,rank_values in ranks.items():

lst = []

groups[major]=lst

# precondition: rank_values已经升序排序

for rank in rank_values:

# 如果跳过了一些组，先全部补充0. post condition: 当前排名所在组的数据存在且是最后一个。

rank_group = rank//group_distance

while len(lst) <= rank_group:

lst.append(0)

lst[rank_group]+=1

while len(lst) <= max_rank//group_distance:

lst.append(0)

for i in range(len(lst)):

lst[i]/=len(rank_values)

lst[i]*=100

indexes = []

fromindex=0

while fromindex <= max_rank:

indexes.append(fromindex)

fromindex+=group_distance

from pprint import pprint as printf

printf(list(groups.values()))

print(indexes)

df = pd.DataFrame(data=groups,index=indexes)

plt.rcParams['font.sans-serif'] = ['SimHei'] # 用来正常显示中文标签

plt.rcParams['axes.unicode_minus'] = False # 用来正常显示负号

cmap = sns.cm.rocket_r

sns.heatmap(df,annot=True,cmap=cmap)

plt.tight_layout()

plt.title(fig_title)

plt.show()