def ptt_zyclass_xktype(self):
# benke zyclass-xktype-count
dt = self.df_class_type_count_bk[['zyclass', 'xk0'] + list(self.field_dict.keys())]
dt = dt.astype({fs: int for fs in list(self.field_dict.keys())+['xk0']})
dt.zyclass = dt.zyclass.apply(lambda x: self.zyclass_name_bk[int(x[0:2])])
dt.loc[:, 'xk_sum'] = sum(dt[fs] for fs in ['xk0']+list(self.field_dict.keys()))
dsum = pd.DataFrame(dt.sum())
dsum.loc['zyclass'] = 'total'
dsum = dsum.unstack().unstack()
# print(dsum)
dt = dt.append(dsum)
align_dict = {fs: 'r' for fs in list(self.field_dict.keys())+['xk0']}
align_dict.update({'zyclass': 'l', 'xk_sum': 'r'})
print(ptt.make_page(dt,
title='xk type count for benke',
align=align_dict))
dtt = pd.DataFrame(self.df_class_type_count_bk.sum()).unstack().unstack()
dtt.zyclass = dtt.zyclass.apply(lambda x: 'total')
dt2 = pd.concat([self.df_class_type_count_bk, dtt])
dt2.zyclass = dt2.zyclass.apply(lambda x: self.zyclass_name_bk[int(x[0:2])] if x != 'total' else x)
dt2 = dt2.astype({fs: int for fs in dt2.columns.values if fs != 'zy' + 'class'})
print(ptt.make_page(dt2,
title='all zy count',
align={fs: 'l' if fs=='zyclass' else 'r' for fs in dt2.columns}))
def find_zhuanye(self,
lowpos=0,
highpos=1000000,
xxfilterlist=('', ),
zyfilterlist=('', )):
# align = dict() if align is None else
if self.dflq is None:
return pd.DataFrame()
xxfilterfun = select_filter(xxfilterlist, '')
zyfilterfun = select_filter(zyfilterlist, '')
df = self.dflq[self.dflq.YXMC.apply(xxfilterfun) & self.dflq.ZYMC.apply(zyfilterfun) & \
(self.dflq.WC >= lowpos) & (self.dflq.WC <= highpos)].\
groupby(['YXDH', 'ZYDH'])[['WC', 'YXMC', 'ZYMC']].max()
if len(df) > 0:
print(
ptt.make_page(df.sort_values('WC'),
''.join(zyfilterlist),
align={
'YXMC': 'l',
'ZYMC': 'l',
'WC': 'r'
}))
else:
print('no record found in pos {}--{} for xx={} zy={}'.format(
lowpos, highpos, xxfilterlist, zyfilterlist))
return # df
def find_tdinfo_from_yxname(self, xxsubstr=('医学', ), kl='wk', cc='bk'):
ffun = closed_filter(xxsubstr)
if ffun is False:
return
# df1, df2, df3 = None, None, None
if cc == 'bk':
print('2016p1---')
df1 = self.td16bk1[self.td16bk1.xx.apply(ffun)][['xx', 'wkpos', 'lkpos']].\
sort_values(by=('lkpos' if kl == 'lk' else 'wkpos'))
print(ptt.make_page(df1, '2016p1'))
print('2016p2---')
df2 = self.td16bk2[self.td16bk2.xx.apply(ffun)][['xx', 'wkpos', 'lkpos']].\
sort_values(by='lkpos' if kl == 'lk' else 'wkpos')
print(ptt.make_page(df2, '2016p2'))
print('2017bk---')
df3 = self.td17bk[self.td17bk.xx.apply(ffun)][['xx', 'wkpos', 'lkpos']].\
sort_values(by='lkpos' if kl == 'lk' else 'wkpos')
print(ptt.make_page(df3, '2017bk', align={'xx': 'l'}))
print('2018bk---')
df3 = self.td18bk[self.td18bk.xx.apply(ffun)][['xx', 'wkpos', 'lkpos']].\
sort_values(by='lkpos' if kl == 'lk' else 'wkpos')
print(ptt.make_page(df3, '2018bk', align={'xx': 'l'}))
else:
# print('2016zk---')
df = self.td16zk[self.td16zk.xx.apply(ffun)][['xx', 'wkpos', 'lkpos']].\
sort_values(by=('lkpos' if kl == 'lk' else 'wkpos'))
print(ptt.make_page(df, title='2016 zk', align={'xx': 'l'}))
# print('2017zk---')
df = self.td17zk[self.td17zk.xx.apply(ffun)][['xx', 'wkpos', 'lkpos']].\
sort_values(by='lkpos' if kl == 'lk' else 'wkpos')
print(ptt.make_page(df, title='2017 zk', align={'xx': 'l'}))
# print('2018zk---')
df = self.td18zk[self.td18zk.xx.apply(ffun)][['xx', 'wkpos', 'lkpos']].\
sort_values(by='lkpos' if kl == 'lk' else 'wkpos')
print(ptt.make_page(df, title='2018 zk', align={'xx': 'l'}))
return # df1, df2, df3
def find_wc_from_score(self, score=500, scope=0, year=18, kl='wk'):
df = None
if year == 18:
if kl.lower() in 'wk, lk':
df = self.fd2018pt
if kl.lower() == 'ys':
df = self.fd2018ystk_zhf
elif year == 19:
if kl.lower() in 'wk, lk':
df = self.fd2019pt
if kl.lower() == 'ys':
df = self.fd2019ystk_zhf
if df is None:
print('no fd data found for kl={} year={}!'.format(kl, year))
print('year in [{}] kl in [{}]'.format('18,19', 'wk,lk, msw, msl'))
return
fdv = df[df.fd.apply(lambda x: score - scope <= x <= score + scope)]
if len(fdv) > 0:
print(
ptt.make_page(fdv,
title=str('score on {} with scope {}'.format(
score, scope))))
else:
print('not founded data for score={}!'.format(score))
def find_tdinfo_from_wc(self,
low,
high,
selecter=(''),
filter=('', ),
kl='wk',
cc='bk',
align=None):
posfield = 'wkpos' if kl == 'wk' else 'lkpos'
align = dict() if align is None else align
if cc == 'bk':
# print('2016pc1---')
df1 = self.get_df_from_pos(self.td16bk1,
lowpos=low,
highpos=high,
posfield=posfield,
selecter=selecter,
filter=filter,
kl=kl)
# print('2016pc2---')
df2 = self.get_df_from_pos(self.td16bk2,
lowpos=low,
highpos=high,
posfield=posfield,
selecter=selecter,
filter=filter,
kl=kl)
# print('2017---')
df3 = self.get_df_from_pos(self.td17bk,
lowpos=low,
highpos=high,
posfield=posfield,
selecter=selecter,
filter=filter,
kl=kl)
df3.loc[:, 'xxh'] = df3.xx.apply(lambda x: str(x)[0:4])
# print('2018---')
df4 = self.get_df_from_pos(self.td18bk,
lowpos=low,
highpos=high,
posfield=posfield,
selecter=selecter,
filter=filter,
kl=kl)
else:
# print('2016zk---')
df1 = self.get_df_from_pos(self.td16zk,
lowpos=low,
highpos=high,
posfield=posfield,
selecter=selecter,
filter=filter,
kl=kl)
df2 = df1.head(0)
# print('2017zk---')
df3 = self.get_df_from_pos(self.td17zk,
lowpos=low,
highpos=high,
posfield=posfield,
selecter=selecter,
filter=filter,
kl=kl)
# print('2018zk---')
df4 = self.get_df_from_pos(self.td18zk,
lowpos=low,
highpos=high,
posfield=posfield,
selecter=selecter,
filter=filter,
kl=kl)
if kl == 'lk':
df1 = df1.rename(columns={'lkjh': 'lkjh16', 'lkpos': 'lkpos16'})
df2 = df2.rename(columns={
'lkjh': 'lkjh16p2',
'lkpos': 'lkpos16p2'
})
df3 = df3.rename(columns={'lkjh': 'lkjh17', 'lkpos': 'lkpos17'})
df4 = df4.rename(columns={'lkjh': 'lkjh18', 'lkpos': 'lkpos18'})
else:
df1 = df1.rename(columns={'wkjh': 'wkjh16', 'wkpos': 'wkpos16'})
df2 = df2.rename(columns={
'wkjh': 'wkjh16p2',
'wkpos': 'wkpos16p2'
})
df3 = df3.rename(columns={'wkjh': 'wkjh17', 'wkpos': 'wkpos17'})
df4 = df4.rename(columns={'wkjh': 'wkjh18', 'wkpos': 'wkpos18'})
for df in [df1, df2, df3, df4]:
df.loc[:, 'xxh'] = df.xx.apply(lambda x: x[0:4])
df1 = df1.drop(labels=['xx'], axis=1)
df2 = df2.drop(labels=['xx'], axis=1)
df3 = df3.drop(labels=['xx'], axis=1)
dfmerge = pd.merge(df4, df3, on='xxh', how='outer')
dfmerge = pd.merge(dfmerge, df1, on='xxh', how='outer') # [outfields]
if df2['xxh'].count() > 0:
dfmerge = pd.merge(dfmerge, df2, on='xxh',
how='outer') # [outfields]
dfmerge = dfmerge.fillna('-1')
f_type = {
k: int
for k in dfmerge.columns if ('jh' in k) or ('pos' in k)
}
dfmerge = dfmerge.astype(dtype=f_type)
dfmerge = dfmerge.drop(labels='xxh', axis=1)
dfmerge = dfmerge.query('xx != "-1"')
print(ptt.make_page(dfmerge, title='data 16-18', align={'xx': 'l'}))
return # dfmerge
def xk_stats(xkfile=data_path_dell+'xk/xk_type_zycount.csv',
plot_pie=False,
ptt_zycount=False,
):
xk_name = ['xk0', 'xk1', 'xk2', 'xk3', 'xk21', 'xk31']
xk_label = ['0', '1', '2', '3', '1/2', '1/3']
xk_subject = ['d', 'h', 'l', 's', 'w', 'z']
xk_sub_cb = cb(xk_subject, 3)
dc = pd.read_csv(data_path_dell+'xk/xk_zyclass_zycount.txt')
dc = dc.fillna(0)
zyclass_name = [x for x in dc.zyclass if x not in ('total','ratio')]
zyclass_name[0] = '00实验基地班'
zyclass_name[-1] = '88中外合作'
# print(zyclass_name)
# read zy class type data
dzy = pd.read_csv(xkfile, dtype={'zyclass': str})
dzy = dzy.fillna(0)
zyfield=list(dzy.columns.values)
zyfield.remove('zyclass')
zy_xk_series = dzy[zyfield].sum()
xk_comb_dict = {}
for xs in xk_sub_cb:
zynum = zy_xk_series['xk0']
for t in zy_xk_series.index:
if '_' not in t:
continue
xktype = t[0:t.find('_')]
xksubs = t[t.find('_')+1:]
if xktype in 'xk1,xk2,xk3':
xss = ''.join(xs)
# print(xksubs, xss)
if xksubs in xss:
print(xss,t,zy_xk_series[t])
zynum += zy_xk_series[t]
elif xktype in 'xk21, xk31':
if len(set(xs) & set(xksubs)) > 0:
print(xs, t, zy_xk_series[t])
zynum += zy_xk_series[t]
xk_comb_dict.update({xs: zynum})
print('km-{} zycount={}'.format(xs, zynum))
dzyp = dzy[dzy.zyclass < '50'] # benke
zy_xk_series_bk = dzyp[zyfield].sum()
field_dict = {
'xk1': [x for x in dzy.columns if 'xk1_' in x],
'xk2': [x for x in dzy.columns if 'xk2_' in x],
'xk3': [x for x in dzy.columns if 'xk3_' in x],
'xk21': [x for x in dzy.columns if 'xk21_' in x],
'xk31': [x for x in dzy.columns if 'xk31_' in x]}
dtemp = dzy.copy()
dtemp2 = dzyp.copy()
for fs in field_dict.keys():
dtemp.loc[:, fs] = sum(dzy[fd] for fd in field_dict[fs])
dtemp2.loc[:, fs] = sum(dzyp[fd] for fd in field_dict[fs])
dzy = dtemp
dzyp = dtemp2
# count for xk_type
xk_count = dzy[xk_name].sum(axis=0)
zyclass_count = dzy[xk_name].sum(axis=1)
type_name = dzy.zyclass
xk_count_bk = dzyp[xk_name].sum()
zyclass_count_bk = dzyp[xk_name].sum(axis=1)
type_name_bk = dzyp.zyclass
xk_count_zk = [x-y for x,y in zip(xk_count, xk_count_bk)]
zyclass_count_zk = [x-y for x,y in zip(zyclass_count, zyclass_count_bk)]
# print(xk_count_zk)
if plot_pie:
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams.update({'font.size': 16})
plt.figure(u'选科专业统计')
plt.subplot(131)
plt.pie(xk_count, labels=xk_label, autopct='%1.2f%%')
plt.title(u'全部选科专业')
plt.subplot(132)
plt.pie(xk_count_bk, labels=xk_label, autopct='%1.2f%%')
plt.title(u'本科选科专业')
plt.subplot(133)
plt.pie(xk_count_zk, labels=xk_label, autopct='%1.2f%%')
plt.title(u'专科选科专业')
if ptt_zycount:
print(ptt.make_page(dzy[['zyclass']+xk_name], title='all zy count'))
print(ptt.make_page(dzyp[['zyclass']+xk_name], title='benke zy count'))
return zy_xk_series, zy_xk_series_bk
def print_xk(self):
print(ptt.make_page(self.df_class_type_count[['zyclass'] + self.xk_type_name], title='all zy count'))
print(ptt.make_page(self.df_class_type_count_bk[['zyclass'] + self.xk_type_name], title='benke zy count'))