如何在sqlalchemist Python的ORM中构建此postgresql查询？

Question

我在ORM SQLalchemyst中构建查询时遇到问题

这是我的SQL：

SELECT DISTINCT eps.expedient_id,  
eps.first_name,
eps.middle_name,
eps.last_name, 
eps.maiden_name,
eps.person_expedient_id, 
eps.full_name,
doc_type,
doc_number,
cud.claimer_id 
FROM phase_2.expedient_person_search eps 
INNER JOIN phase_2.claimer_user_data cud ON eps.expedient_id = cud.expedient_id 
WHERE eps.first_name = data.first_name AND doc_type = data.doc_type 
ORDER BY order_criteria DESC sorting_order, expedient_id   
LIMIT 10 OFFSET 1

我已经尝试过这种方式，但是在运行查询时总是会收到错误消息：

Expedients.query.distinct(Expedients.expedient_id).order_by(Expedients.doc_type.desc()).filter_by(**data).offset(page).limit(perPage).all()

并显示此错误：

Error: (psycopg2.ProgrammingError) SELECT DISTINCT ON expressions must match initial ORDER BY expressions

有人可以告诉我我在做什么错吗？

我的模特：

class Expedients(db.Model):
    __tablename__ = "expedient_person_search"
    __table_args__ = {'schema':'phase_2'}
    expedient_id = db.Column(db.Integer, primary_key=True)
    radicate_number = db.Column(db.String(40))
    doc_type = db.Column(db.String(10))
    doc_number = db.Column(db.String(20),db.ForeignKey('phase_2.claimer_user_data.document_number'))
    first_name = db.Column(db.String(128))
    middle_name = db.Column(db.String(128))
    last_name = db.Column(db.String(128))
    maiden_name = db.Column(db.String(128))
    person_expedient_id=db.Column(db.Integer)
    full_name = db.Column(db.String(128))

编辑22/03/2019

用户模型：

class User(db.Model):
    __tablename__ = "claimer_user_data"
    __table_args__ = {'schema':'phase_2'}
    claimer_id = db.Column(db.Integer, primary_key=True, autoincrement=True)
    expedient_id = db.Column(db.Integer)
    first_name = db.Column(db.String(128))
    last_name = db.Column(db.String(128))
    document_type = db.Column(db.String(10))
    document_number = db.Column(db.String(20))
    username = db.Column(db.String(250))
    password_hash = db.Column(db.String(100))
    expedients = db.relationship("Expedients")

（具有更多列）

这是mi（** data）数据模型json：

{
  "doc_type": "string",
  "doc_number": "string",
  "first_name": "string",
  "middle_name": "string",
  "last_name": "string",
  "maiden_name": "string",
  "radicate_number": "string",
  "person_expedient_id": "string",
  "page": 0,
  "perPage": 0,
  "orderCriteria": "string",
  "sortingOrder": "string"
}