I'm trying to develop custom human ressources evaluation dashboards using a custom module.

I took the existing hr_evaluation dashboard as a start, and try to do such developpement.

A- Python side : 

So I found that I have to create a report model, "hr_evaluation_report" that takes fields from models involved on the dashboard (here the hr_evaluation_evaluation and hr_evaluation_interview models), and if needed some fields used to display computed results on the dashboard (here the delay_date and overpass_delay), and this using depends, as the following code demonstrates : 

hr_evaluation_report.py 

In this file I found also a init method.

I need to know more about this init method :

1- Is it mandatory ?

2- Is it serving only for performing a sql request to make data available for the reporting model and then initialize it?

3- I see that it reacts on views displaying using tools.drop_view_if_exists, so what are all the purposes for using this method ?

B- View side :

On the hr_evaluation_view.xml 

We can see that :

1- Initializes the graph view

2- Making Filters and group by availables

3- Adding dashboard menuitem, and action allowing to open the corresponding view.

But I can't see :

How to customize the measure selectio box ?

Image showing the measures selection box 

And a final question about this button sequence :

Image showing the sequence button 

Does this buttons sequence come by default (using inheritance maybe) ? 

Every suggestion, tutorials, additions, hints will be welcome. 

------------------------------------------------- EDIT --------------------------------------------------------------------------------------------------------------------

@Axel Mendoza : 

I developped my custom module for reporting, without the init method, and I set _auto=True, so Odoo will create the sql table for me on the database. Here is the main module files : 

1. hr_evaluation_report_extension.py

On this new created model, I want to get the necessary fields needed to retrieve data, the purpose is to have the goals reaching percentage per employee, and having the ability to group by employee, department, job, ... , So the 'reach' field is the one measurable, because it's the only one with float type.

So I needed to put 3 models on the depends list, but there is no relationnal fields between the three, Is it necessary to have data consistency ? because on the example I followed, when using the init method, they did a LEFT JOIN to have data, what did I need in case I don't use the init method ?

from import openerp.osv fields, osv


class hr_evaluation_report_goals(osv.Model):
    _name = "hr.evaluation.report.goals"
    "Goals reaching Statistics"
    _description = _auto = True

    _columns = {
        'create_date': fields.datetime('Create Date', readonly=True),

        'plan_id': fields.many2one('hr_evaluation.plan','Plan' , readonly=True),

        'job_id': fields.many2one('hr.job','Job Title ),

        'department_id': fields.many2one('hr.department', 'Department'),

        'statut': fields.char('Statut', size=32),

        'name_related': fields.related('resource_id','name', type='char', string='Name', readonly=True, store=True),

        'reach': fields.float('Atteinte objectif global'),

        'state': fields.selection([
            ('draft''Draft', ),
            ('wait''Plan In Progress', ),
            ('progress''Final Validation', ),
            ('done''Done', ),
            ('cancel''Cancelled', ),
        ], 'Status', readonly=True),
    }
    _depends = {
        'hr.employee': ['name_related','job_id','department_id','statut'],

        'survey.user_input': ['reach'],

        'hr_evaluation.evaluation': [
            'create_date','plan_id','state'],
    }

The problem is there is no data on the sql table created by Odoo :

Image showing the sql request result

2. hr_evaluation_report_extension_view.xml

On the view side I put on the graph tag the name_related (employee name) field on the rows, and create_date (employee evaluation create date) on the col, and the field reach as measure.

besides, I created the search view for filters and group by.

<?xml version="1.0" encoding="utf-8"?>
<openerp>
    <data>

        <record id="view_goals_reaching_graph" model="ir.ui.view">
                <field name="name">hr.goals.reaching</field>
                <field name="model">hr.evaluation.report.goals</field>
                <field name="arch" type="xml">
                    <graph string="Goals reaching analysis" type="pivot" stacked="True">
                         <field name="name_related" type="row"/>
                         <field name="create_date" type="col"/>
                         <field name="reach" type="measure"/>
                     </graph>
                </field>
        </record>

        <record id="view_goals_reaching_search" model="ir.ui.view">
            <field name="name">hr.goals.search</field>
            <field name="model">hr.evaluation.report.goals</field>
            <field name="arch" type="xml">
                <search string="Pourcentage atteinte globale des objectifs">
                    <filter string="En cours" domain="[('state', '=' ,'wait')]" help = "In progress Evaluations"/>
                    <filter string="Attente appréciation DRH" domain="[('state','=','progress')]" help = "Final Validation Evaluations"/>
                    <filter string="Terminé" domain="[('state','=','done')]"/>
                    <field name="name_related"/>
                    <field name="plan_id"/>
                    <group expand="0" string="Extended Filters...">
                        <field name="state"/>
                        <field name="create_date"/>
                        <field name="department_id"/>
                        <field name="job_id"/>
                        <field name="statut"/>
                    </group>
                    <group expand="1" string="Group By">
                        <filter string="Employé" name="employee" context="{'group_by':'name_related'}"/>
                        <filter string="Campagne" context="{'group_by':'plan_id'}"/>
                        <filter string="Direction" context="{'group_by':'department_id'}"/>
                        <filter string="Fonction" context="{'group_by':'job_id'}"/>
                        <filter string="statut" context="{'group_by':'statut'}"/>
                        <separator/>
                        <filter string="Month" context="{'group_by':'create_date:month'}" help="Mois de création"/>
                    </group>
                </search>
            </field>
        </record>

        <record id="action_evaluation_goals_reaching" model="ir.actions.act_window">
            <field name="name">Atteinte des objectifs</field>
            <field name="res_model">hr.evaluation.report.goals</field>
            <field name="view_type">form</field>
            <field name="view_mode">graph</field>
            <field name="search_view_id" ref="view_goals_reaching_search"/>
        </record>

        <menuitem action="action_evaluation_goals_reaching" id="menu_evaluation_goals_reaching" parent="hr.menu_hr_reporting" sequence="3" groups="base.group_hr_manager"/>

    </data>
</openerp>

-------------------------------------------------------------------EDIT 2-------------------------------------------------------------------------------------------------

Here after the report model : 

class hr_evaluation_report_goals(osv.Model):
    _name = "hr.evaluation.report.goals"
    _description = "Goals reaching"
    _auto = False

    _columns = {
        'evaluation_plan_id': fields.many2one('hr_evaluation.plan , 'Appraisal Plan' ),

        'job_id': fields.many2one('hr.job', 'Job Title'),

        'department_id': fields.many2one('hr.department', 'Department'),

        'statut': fields.char('Statut', size=32),

        'name_related': fields.related('resource_id', 'name' , type='char', string='Name', readonly=True, store=True),

        'reach': fields.float('Atteinte objectif global'),

    }
    _depends = {
        'hr.employee': ['id','name_related','job_id','department_id','statut','evaluation_plan_id'],

        'survey.user_input': ['reach']

    }

Here is my init method : 

def init(self, cr):
    tools.drop_view_if_exists(cr, 'hr_evaluation_report_goals')
    cr.execute("""
        create or replace view hr_evaluation_report_goals as (
             select
                 min(e.id) as id,
                 e.name_related,
                 e.evaluation_plan_id,
                 e.department_id,
                 e.job_id,
                 e.statut,
                 s.reach
                 from
             survey_user_input s
            INNER JOIN
                 hr_employee e on (e.id=s.employee_id)
             GROUP BY
                 e.name_related,
                 e.evaluation_plan_id,
                 e.department_id,
                 e.job_id,
                 e.statut
        )
    """)

I get this error :

ProgrammingError: column "s.reach" must appear in the GROUP BY clause or be used in an aggregate function

LINE 10: s.reach

How to store my goals reaching value in the sql view, if it's not syntaxclty correct ?