def statistic_rank_dimension(df, groups, na_field, fields, is_sa=True, is_cube=False, sort_field='-avg_salary',
need_total=False):
"""
排名
"""
na_field = [na_field] if na_field else []
fields = [fields] if fields else []
all_fields = groups + na_field + fields
for field in all_fields:
df = df.filter(df[field].isNotNull())
if is_cube:
md_df = statistic_cube(df, groups, na_field, fields, is_sa)
else:
md_df = statistic_groups(df, groups, na_field, fields, is_sa)
if need_total:
md_df = add_total(md_df, *(groups + na_field))
if fields:
md_df = add_rank(md_df, *(groups + na_field), sort_field=sort_field)
else:
md_df = add_rank(md_df, *(groups[1:] + na_field), sort_field=sort_field)
return md_df
def statistic_salary_dimension(df,
groups,
na_field,
fields,
is_sa=True,
is_cube=False,
is_age=False,
need_alias=False):
"""
薪资分析
:param df:
:return:
"""
na_field = [na_field] if na_field else []
fields = fields.split(',') if fields else []
all_fields = groups + na_field + fields
for f in all_fields:
df = df.filter(df[f].isNotNull())
if is_age:
df = filter_age(df).withColumn("age", F.udf(str)(df.age))
if is_cube:
md_df = statistic_cube(df, groups, na_field, fields, is_sa)
else:
md_df = statistic_groups(df, groups, na_field, fields, is_sa)
if need_alias:
md_df = md_df.join(get_position_industry(df), "position_name")
md_df = md_df.join(get_position_alias(df), "position_name")
return md_df
def statistic_position_address_change(df,
groups,
na_field,
fields,
is_sa=False):
"""
职位延续性分析
:param df:
:return:
"""
na_field = [na_field] if na_field else []
fields = [fields] if fields else []
all_fields = groups + na_field + fields
df = df.select(all_fields[0], all_fields[1], "resume_id", "work_index")
df = next_same(df, groups[0])
md_df = statistic_cube(df, groups, na_field, fields, is_sa)
md_df = add_rank(md_df, *(groups[1:] + na_field), sort_field='-person_num')
return md_df
def statistic_number_dimension(df,
groups,
na_field,
fields,
is_sa=False,
is_cube=False):
"""
人数（权重）分析
:param df:
:return:
"""
na_field = [na_field] if na_field else []
fields = fields.split(',') if fields else []
all_fields = groups + na_field + fields
for f in all_fields:
df = df.filter(df[f].isNotNull())
if is_cube:
md_df = statistic_cube(df, groups, na_field, fields, is_sa)
else:
md_df = statistic_groups(df, groups, na_field, fields, is_sa)
return md_df