def analyze_admission(self, year=None):
connect = DatabaseHelper()
data = connect.get_all_admission(year)
value = {}
if data['value']:
df = pd.DataFrame(data['value'])
# real data
# deparment = connect.get_department()
# deparment = pd.io.json.json_normalize(deparment['value'], max_level=0)
# # branch_data = analyze_helper.set_branch(branch['value'])
# deparment_data = deparment[['dept_id','dept_name']]
# deparment_data.set_index('dept_id',inplace=True)
deparment = connect.get_department_ds()
deparment_data = pd.DataFrame(deparment['value'])
deparment_data = deparment_data[['dept_id', 'dept_name']]
deparment_data.set_index('dept_id', inplace=True)
status_data = analyze_helper.set_fullname(
connect.get_status_list())
channel_data = analyze_helper.set_fullname(
connect.get_admission_channel())
channel_sample = self.split_channel(channel_data)
###check duplicate channel anme
dupli_channel = channel_sample.channel_name.duplicated(keep=False)
channel_sample.loc[dupli_channel, 'channel_name'] = channel_sample.loc[dupli_channel, 'channel_name'] + ' (' + \
channel_sample['channel_round'] + ')'
channel_sample_for_dict = channel_sample[['channel_name']]
round_data = channel_sample[['round_id', 'round_name']]
school = analyze_helper.set_fullname(connect.get_school_lis())
deparment_dict = analyze_helper.set_dict(deparment_data.index,
deparment_data.dept_name)
school_dict = analyze_helper.set_dict(school.index,
school.school_title)
status_dic = analyze_helper.set_dict(status_data.index,
status_data.status_title)
channel_dic = analyze_helper.set_dict(
channel_sample_for_dict.index,
channel_sample_for_dict.channel_name)
data_split_now = df.copy()
data_not_year = df.copy()
if year:
data_split_now = df.loc[df['admission_year'] == int(year)]
round_list = channel_sample['round_id'].unique().tolist()
data_not_none = data_split_now.dropna()
gpa_by_branch = []
response_round = False
for i in round_list:
gpa_by_count = {}
channel_list = channel_sample[channel_sample['round_id'] == i]
list_channel = channel_list.index.tolist()
if (not data_not_none.empty):
count_by_branch = data_not_none['channel_id'].isin(
list_channel)
count_by_branch = data_not_none[count_by_branch]
count_by_branch = count_by_branch.groupby([
'channel_id', 'dept_id'
])['current_gpax'].mean().unstack(fill_value=0)
count_by_branch = count_by_branch.round(2)
count_by_branch_check_branch = analyze_helper.check_list_column(
deparment_data.index, count_by_branch)
count_by_branch_check_channel = analyze_helper.check_list(
channel_list.index, count_by_branch_check_branch)
check_by_round_channel = analyze_helper.set_fullname_column(
deparment_dict, count_by_branch_check_branch)
check_by_round_channel = analyze_helper.set_fullname_index(
channel_dic, check_by_round_channel)
gpa_by_count[
'gpa_by_branch'] = check_by_round_channel.to_dict(
'index')
gpa_by_branch.append(gpa_by_count)
response_round = True
count_channel = data_split_now.groupby('channel_id').size()
count_channel_check_channel = analyze_helper.check_list(
channel_sample.index, count_channel)
count_channel_check_channel = analyze_helper.set_fullname_index(
channel_dic, count_channel_check_channel)
count_school = data_split_now.school_id.value_counts()
sort_count_school = count_school.sort_values(
ascending=False).head()
sort_count_school_fullname = analyze_helper.set_fullname_index(
school_dict, sort_count_school)
count_by_status = data_split_now.groupby(
['channel_id', 'status_id']).size().unstack(fill_value=0)
count_by_status_check_status = analyze_helper.check_list_column(
status_data.index, count_by_status)
count_by_status_check_channel = analyze_helper.check_list(
channel_sample.index, count_by_status_check_status)
count_by_status_fullname = analyze_helper.set_fullname_column(
status_dic, count_by_status_check_channel)
count_by_status_fullname = analyze_helper.set_fullname_index(
channel_dic, count_by_status_fullname)
# used year and branch but used year and year-1
year_select = []
if year:
year_select.append(int(year))
year_select.append(int(year) - 1)
else:
max_year = data_not_year.admission_year.max()
year_select.append(max_year)
year_select.append(max_year - 1)
data_compare = data_not_year[data_not_year['admission_year'].isin(
year_select)]
compare_year = data_compare.groupby(
['channel_id', 'admission_year']).size().unstack(fill_value=0)
if compare_year.empty:
channel_name = channel_sample['channel_name'].unique().tolist()
compare_year_success = pd.DataFrame(0,
index=np.arange(
len(channel_name)),
columns=year_select)
compare_year_success['channel_name'] = channel_name
compare_year_success.set_index('channel_name', inplace=True)
else:
compare_year_check_channel = analyze_helper.check_list(
channel_sample.index, compare_year)
compare_year_success = analyze_helper.check_list_column(
year_select, compare_year_check_channel)
compare_year_success = analyze_helper.set_fullname_index(
channel_dic, compare_year_success)
value = {
'count_channel': count_channel_check_channel.to_dict(),
'count_by_branch': dict(zip(round_list, gpa_by_branch)),
'count_by_school': [sort_count_school_fullname.to_dict()],
'compare_year': [compare_year_success.to_dict('index')],
'count_by_status': [count_by_status_fullname.to_dict('index')],
'response_round': response_round
}
response = True
message = "Analyze Successfully"
else:
response = False
message = "Don't have Data"
return inner_res_helper.make_inner_response(response=response,
message=message,
value=value)
def analyze_student_status(self, year=None):
connect = DatabaseHelper()
data = connect.get_all_admission(year)
value = {}
if data['value']:
df = pd.DataFrame(data['value'])
df = df[df['admission_year'] == int(year)]
if not df.empty:
status_data = analyze_helper.set_fullname(
connect.get_status_list())
status_dic = analyze_helper.set_dict(status_data.index,
status_data.status_title)
channel_data = analyze_helper.set_fullname(
connect.get_admission_channel())
channel_sample = self.split_channel(channel_data)
dupli_channel = channel_sample.channel_name.duplicated(
keep=False)
channel_sample.loc[dupli_channel, 'channel_name'] = channel_sample.loc[dupli_channel, 'channel_name'] + ' (' + \
channel_sample['channel_round'] + ')'
channel_dict = analyze_helper.set_dict(
channel_sample.index, channel_sample.channel_name)
branch = connect.get_branch()
branch_data = analyze_helper.set_branch(branch['value'])
branch_dict = analyze_helper.set_dict(branch_data.index,
branch_data.branch_name)
group_brance = df.groupby(['channel_id', 'branch_id'
]).size().unstack(fill_value=0)
group_brance = analyze_helper.check_list_column(
branch_data.index, group_brance)
group_brance = analyze_helper.check_list(
channel_data.index, group_brance)
group_brance = analyze_helper.set_fullname_column(
branch_dict, group_brance)
group_brance = analyze_helper.set_fullname_index(
channel_dict, group_brance)
branch_list = group_brance.columns.tolist()
channel_id_list = channel_data.index.tolist()
table_count = []
for c_id in channel_id_list:
by_channel = {}
data = df[df['channel_id'] == c_id]
if not data.empty:
count = len(data)
max_data = data.branch_id.value_counts().max()
min_data = data.branch_id.value_counts().min()
else:
count = 0
max_data = 0
min_data = 0
by_channel['channel'] = channel_dict[c_id]
by_channel['count'] = str(count)
by_channel['max_data'] = str(max_data)
by_channel['min_data'] = str(min_data)
table_count.append(by_channel)
all_student = len(df)
channel_count = df.channel_id.value_counts()
group = df[(df['status_id'] == 2) | (df['status_id'] == 3)]
group = group.groupby(['channel_id', 'status_id'
]).size().unstack(fill_value=0)
if group.empty:
group = pd.DataFrame(0,
index=np.arange(len(channel_count)),
columns=[2, 3])
group['channel'] = channel_count.index
group.set_index('channel', inplace=True)
group = group.rename(columns={2: "probation", 3: "drop"})
list_name = group.columns.tolist()
group = pd.merge(channel_count,
group,
left_index=True,
right_index=True,
how='inner')
group.rename(columns={group.columns[0]: "all"}, inplace=True)
all_admission = group['all'].sum()
for name in list_name:
group['per_Type_' + str(name)] = group.apply(
lambda row: (row[name] / row['all']) * 100, axis=1)
group['per_Stu_' + str(name)] = group.apply(
lambda row: (row[name] / all_student) * 100, axis=1)
group['per_all_student'] = (group['all'] / all_admission) * 100
group = group.round(2).sort_index()
group_check_index = analyze_helper.check_list(
channel_data.index, group)
group_fullname = analyze_helper.set_fullname_index(
channel_dict, group_check_index)
value = {
'branch': branch_list,
'count_by_brance': group_brance.to_dict('index'),
'all_student': str(all_student),
'table': group_check_index.to_dict('index'),
'table_count': table_count,
}
response = True
message = "Don't have Data"
else:
value = {}
response = False
message = "Don't have Data"
else:
value = {}
response = False
message = "Don't have Data"
return inner_res_helper.make_inner_response(response=response,
message=message,
value=value)