2023 New Features

Master Data Management

User Interactions

Add / Edit Administration Attribute

• Step 1: Click the "Add Attribute" button.
• Step 2: Fill in the attribute name and select the type (e.g., "Value","Option", "Multiple Option", "Aggregate").
• Step 3: If the attribute type is "Option,Multiple Option or Aggregate" click the "+" button to add more options.
• Step 4: Click "Submit" to save.
• Step 5: Success alert message appears and return to Administration Attribute list

API: administration-attribute-crud

Add / Edit Administration

• Step 1: Click "Add New Administration" or select an administrative area to edit.
• Step 2: Select Level Name.
• Step 3: Select the parent administration using a cascading drop-down.
• Step 4: Fill in administration details (name, parent, and code).
• Step 5: Fill in attributes and their values.
  
  • For Value type: Input Number
  • For Option and Multiple Option type: Drop-down option
  • For Aggregate: It will shows table with 2 columns, the columns are: name, value
    • Name: the dissagregation name
    • Value: Input Number
• Step 6: Click "Submit" to save.
• Step 7: Confirmation message appears.
• Step 8: Options to return to administration list.

API: administration-crud

Administration / Entity Attribute Types

Option & Multiple Option Values

Use Case

We have a dataset that contains categorical information about the types of land use for various regions. This data will be utilized to classify and analyze land use patterns at the county level.

Feature

To achieve this, we will need to define option values for an attribute. In this scenario, the workflow is as follows:

Define Attribute

Define Option Values

Upload Data for Counties

In this case, we define the "Option Values" for the "Land Use Type" attribute, allowing us to categorize land use patterns at the county level. The actual data for individual counties is then uploaded using the defined options.

Single Numeric Values

Use Case

We possess household counts from the 2019 census that correspond to the RTMIS administrative list at the sub-county level. This data can be employed to compute the household coverage per county, which is calculated as (# of households in that sub-county in RTMIS / # from the census).

Feature

To achieve this, we need to store the population value for individual sub-counties as part of their attributes. In this scenario, the workflow is as follows:

Define Attribute

• Attribute Name: Census HH Count
• Attribute Code: <Unique Identifier>Census_HH_Count
• Type: Single Numeric Value
• Administration Level: Sub-County

Upload Data for Individual Sub-Counties

Sub-County	Attribute Code	Value
CHANGAMWE	Census_HH_Count	46,614
JOMVU	Census_HH_Count	53,472

In this case, the values for the county level will be automatically aggregated.

Disaggregated Numeric Values

Use Case

We aim to import data from the CLTS platform or the census regarding the count of different types of toilets, and we have a match at the sub-county level. This data will serve as baseline values for visualization.

Feature

For this use case, we need to store disaggregated values for an attribute. To do so, we will:

Define the Attribute

• Attribute Name: Census HH Toilet Count
• Attribute Code: <Unique Identifier>Census_HH_Toilet_Count
• Type: Disaggregated Numeric Values
• Disaggregation: “Improved”, “Unimproved”
• Administration Level: Sub-County

Upload Data for Individual Sub-Counties

Sub-County	Attribute Code	Disaggregation	Value
CHANGAMWE	Census_HH_Toilet_Count	Improved	305,927
CHANGAMWE	Census_HH_Toilet_Count	Unimproved	70,367

Bulk Upload

As an administrator of the system, the ability to efficiently manage and update administrative data is crucial. To facilitate this, a feature is needed that allows for the bulk uploading of administrative data through a CSV file. This CSV file format is generated based on the administration attribute table, ensuring that it includes all necessary columns for both administrative levels and their respective attributes.

The CSV template, derived from the administration attribute table, will contain columns representing various administrative levels (such as National, County, Sub-County, Ward, and Village) along with their respective IDs. Additionally, it will include columns for different attributes associated with each administrative unit, as defined in the administration attribute table.

Acceptance Criteria

CSV File Format and Structure

• The system should accept CSV files for bulk upload.
• The CSV file must include columns for different administrative levels (e.g., National, County, Sub-County, Ward, Village).
• Each administrative level column in the CSV file must be filled to ensure proper hierarchical placement.
• Columns for administrative codes and attributes are included but are optional to fill.

Optional Codes and Attributes

• While the administrative codes and attribute columns are provided in the CSV template, filling them is optional.
• The system should be able to process the CSV file and update the administration data correctly, even if some or all of the code and attribute columns are left blank.

Data Validation and Integrity

• The system should validate the CSV file to ensure that all required administrative level columns are filled.
• The system should handle empty optional fields (codes and attributes) gracefully without causing errors.
• Any discrepancies or format errors in the CSV file should be reported back to the user for correction via email.
• The system should process the CSV file efficiently, updating existing records and adding new ones as necessary.
• The process should be optimized to handle large datasets without significant performance issues.

User Feedback and Error Handling

• The user should receive clear feedback on the progress of the upload, including confirmation via email once the upload is complete.
• The system should provide detailed error messages or guidance in case of upload failures or data inconsistencies.

Access Control

• Access to the bulk upload feature should be restricted to authorized users only, with Super Admin role.

Notes

• This user story focuses on streamlining the process of updating administrative data, making it more efficient and less prone to errors than manual entry.
• The integration with the existing Levels model ensures consistency and proper categorization of the administrative units within the system's framework.

Database Overview

Entities Table

pos	table	column	null	dtype	len	default
1	Entities	id		Integer
2	Entities	name		Text

Entity Data Table

pos	table	column	null	dtype	len	default
1	Entity Data	id		Integer
2	Entity Data	entity_id		Integer
3	Entity Data	name		Text
4	Entity Data	administration_id		Integer

Entity Attributes

pos	table	column	null	dtype	len	default
1	Entity Attributes	id		Integer
2	Entity Attributes	entity_id		Integer
3	Entity Attributes	name		Text

Entity Attributes Options

pos	table	column	null	dtype	len	default
1	Entity Attributes Options	id		Integer
2	Entity Attributes Options	entity_attribute_id		Integer
3	Entity Attributes Options	name		Text

Entity Values

pos	table	column	null	dtype	len	default
1	Entity Values	id		Integer
2	Entity Values	entity_data_id		Integer
3	Entity Values	entity_attribute_id		Integer
4	Entity Values	value		Text

Administration Table

pos	table	column	null	dtype	len	default
1	administrator	id	NO	bigint		administrator_id_seq
2	administrator	code	YES	character varying	255
3	administrator	name	NO	text
4	administrator	level_id	NO	bigint
5	administrator	parent_id	YES	bigint
6	administrator	path	YES	text

Administration Attributes

pos	table	column	null	dtype	len	default
1	Administration Attributes	id		Integer
2	Administration Attributes	level_id		Integer
3	Administration Attribute	code		Text		Unique (Auto-Generated)
4	Administration Attributes	Type		Enum (Number, Option, Aggregate)
5	Administration Attributes	name		Text

Administration Attributes Options

pos	table	column	null	dtype	len	default
1	Administration Attributes Options	id		Integer
2	Administration Attributes Options	administration_attributes_id		Integer
3	Administration Attributes Options	name		Text

Administration Values

pos	table	column	null	dtype	len	default
1	Administration Values	id		Integer
2	Administration Values	administration_id		Integer
3	Administration Values	administration_attributes_id	Integer
4	Administration Values	value		Integer
5	Administrative Values	option		Text

Rules:

• Attribute Type: Numeric
  • value: NOT NULL
  • option: NULL
• Attribute Type: Option
  • value: NULL
  • option: NOT NULL
• Attribute Type: Aggregate
  • value: NOT NULL
  • option: NOT NULL

Validation for Option Type

• If parent has a value for a particular administration_attributes_id, then invalidate the children input.
• If children have a value for a particular administration_attributes_id, then override the children value.

Materialized View for Aggregation

Visualization Query

id	type	name	attribute	option	value
1	administration	Bantul	Water Points Type	Dugwell	1
2	entity	Bantul School	Type of school	Highschool	1

API Endpoints

Administration Endpoints

Administration CRUD (POST & PUT)

{
  "parent_id": 1,
  "name": "Village A",
  "code": "VA",
  "attributes": [{
      "attribute":1,
      "value": 200,
    },{
      "attribute":2,
      "value": "Rural",
    },{
      "attribute":3,
      "value": ["School","Health Facilities"],
    },{
      "attribute":4,
      "value": {"Improved": 100,"Unimproved": 200},
    }
  ]
}

Administration Detail (GET)

{
  "id": 2,
  "name": "Tiati",
  "code": "BT",
  "parent": {
    "id": 1,
    "name": "Baringo",
    "code": "B"
  },
  "level": {
    "id": 1,
    "name": "Sub-county"
  },
  "childrens": [{
    "id": 2,
    "name": "Tiati",
    "code": "BT"
  }],
  "attributes": [{
      "attribute":1,
      "type": "value",
      "value": 200,
    },{
      "attribute":2,
      "type": "option",
      "value": "Rural",
    },{
      "attribute":3,
      "type": "multiple_option",
      "value": ["School","Health Facilities"],
    },{
      "attribute":4,
      "type": "aggregate",
      "value": {"Improved": 100,"Unimproved": 200},
    }
  ]
}

Administration List (GET)

Query Parameters (for filtering data):

• parent (only show data that has same parent id, so the parent itself should not be included)
  
• search (search keyword: by name or code)
• level
• Rules:
  • Always filter parent_id = null (Kenya) by default

{
  "current": "self.page.number",
  "total": "self.page.paginator.count",
  "total_page": "self.page.paginator.num_pages",
  "data":[
    {
      "id": 2,
      "name": "Tiati",
      "code": "BT",
      "parent": {
        "id": 1,
        "name": "Baringo",
      },
      "level": {
        "id": 1,
        "name": "Sub-county"
      }
    }
]}

Administration Attributes CRUD (POST & PUT)

{
  "name": "Population",
  "type": "value",
  "options": []
}

Administration Attributes (GET)

[{
  "id": 1,
  "name": "Population",
  "type": "value",
  "options": []
},{
  "id": 2,
  "name": "Wheter Urban or Rural",
  "type": "option",
  "options": ["Rural","Urban"]
},{
  "id": 3,
  "name": "HCF and School Availability",
  "type": "multiple_option",
  "options": ["School","Health Care Facilities"]
},{
  "id": 4,
  "name": "JMP Status",
  "type": "aggregate",
  "options": ["Improved","Unimproved"]
}]

Database Seeder

Administration Seeder

In the updated approach for seeding initial administration data, the shift from using TopoJSON to Excel file format is being implemented. While TopoJSON has been the format of choice, particularly for its geospatial data capabilities which are essential for visualization purposes, the move to Excel is driven by the need for a more flexible and user-friendly data input method.

However, this transition introduces potential challenges in maintaining consistency between the Excel-based administration data and the TopoJSON used for visualization. The inherent differences in data structure and handling between these two formats could lead to discrepancies, impacting the overall data integrity and coherence in the system. This change necessitates a careful consideration of strategies to ensure that the data remains consistent and reliable across both formats.

Key Considerations

• Data Format and Consistency: The shift to Excel might introduce inconsistencies with the TopoJSON format, especially in terms of data structure and geospatial properties.
• Data Validation: Robust validation is essential to mitigate errors common in Excel files.
• Import Complexity: Managing complex Excel structures requires additional parsing mechanisms.
• Scalability and Performance: Excel's performance with large datasets and memory usage should be monitored.
• Security and Integrity: Increased risk of data tampering in Excel files, and challenges in version control.
• Automation and Workflow Integration: Adapting automation processes to accommodate Excel's format variations.
• User-Provided Data: Dependence on external data updates necessitates clear handling policies.

Excel File Structure for Seeder

File Naming Convention

• Each Excel file represents a county.
• File names follow the format: <county_id>-<county_name>.xlsx
• Example: 101-Nairobi.xlsx, 102-Mombasa.xlsx

File Content Structure

Each file contains details of sub-counties and wards within the respective county.

Sub-County_ID	Sub-County	Ward_ID	Ward
201	Westlands	301	XYZ
201	Westlands	302	ABC
...	...	...	...

Seeder Adaptation

• Hard-coded National Level: The national level, Kenya, should be hard-coded in the seeder.
• Dynamic County Processing: The seeder dynamically processes each county file, creating or updating records for sub-counties and wards.
• File Processing Logic: The seeder reads the file name to determine the county and iterates through each row to seed data for sub-counties and wards.

Administration Attribute Seeder

Assumptions

• Administration IDs are available and consistent.
• The attributes are stored in an Excel file, with a structure that includes administration IDs and their corresponding attributes.

Example Excel File Structure

Admin_ID	Attribute1	Attribute2	...
1	Value1	Value2	...
2	Value1	Value2	...
...	...	...	...

Seeder Script

import pandas as pd
from your_app.models import Administration, AdministrationAttribute

class AdministrationAttributeSeeder:
    def __init__(self, file_path):
        self.file_path = file_path

    def run(self):
        # Load data from Excel file
        df = pd.read_excel(self.file_path)

        # Iterate through each row in the DataFrame
        for index, row in df.iterrows():
            admin_id = row['Admin_ID']
            # Retrieve the corresponding Administration object
            administration = Administration.objects.get(id=admin_id)

            # Create or update AdministrationAttribute
            for attr in row.index[1:]:  # Skipping the first column (Admin_ID)
                attribute_value = row[attr]
                AdministrationAttribute.objects.update_or_create(
                    administration=administration,
                    attribute_name=attr,
                    defaults={'attribute_value': attribute_value}
                )

        print("Administration attributes seeding completed.")

# Usage
seeder = AdministrationAttributeSeeder('path_to_your_excel_file.xlsx')
seeder.run()

Note:

1. File Path: Replace 'path_to_your_excel_file.xlsx' with the actual path to the Excel file containing the administration attributes, the excel files will be safely stored in backend/source.
2. Model Structure: This script assumes the existence of Administration and AdministrationAttribute models. Adjust the script according to your actual model names and structures.
3. update_or_create: This method is used to either update an existing attribute or create a new one if it doesn't exist.
4. Error Handling: Add appropriate error handling to manage cases where the administration ID is not found or the file cannot be read.