Build the Insights Panel

[mvmaltitz]

Workflow

See proposed workflow here:
https://app.diagrams.net/#G1_bCSuTSAArfclRQ0zld_ZU915tDF0i1O#%7B%22pageId%22%3A%228YMj8Y18EC_aEH8-hYeU%22%7D

Wireframe

Enablement

This tab enables the user to obtain further refined insights from the score rasters produced in the previous tabs based on discrete descriptive classes. Population data can also be combined with the discrete score output dataset, allowing for more nuanced insights that account for both the level of enablement and population. Additionally, this tab provides the option to aggregate results based on specific boundaries of interest.

Step 1: Classify into discrete classes

Enablement Score Input layer (required): The dimension or final weighted aggregated score layer of interest (.tif). Scores are grouped into five discrete classes as defined in the 'About' Tab.

Output (Level of Enablment): The enablement score input layer classified into five discrete level of enablement classes.

Population Input layer (required): A population count layer (.tif). This should be represented as the count of women in each raster cell.

Output (Relative Population Count): The population input layer classified into three discrete classes based on the lower quartile range, interquartile range, and upper quartile range of data to identify areas of relatively low, medium, and high population per region.

Step 2: Combine score and population classifications

Level of Enablment Input Layer: The level of enablement output raster layer produced in step 1.

Relative Population Count Input Layer: The relative population count raster layer produced in step 1.

Output (Combine Classification): A single raster layer which combines the five levels of enablement classes and three population classes. This layer contains 15 classes defined in the ‘About’ tab. This output covers the entire country or region of interest.

Step 3: Aggregation

Combine Classification Input Layer: The combine classification output raster layer produced in step 2.

Aggregation Input Layer: A polygon layer representing boundaries of interest for aggregation (e.g. municipal boundary layer)

Output (Aggregation): A polygon layer showing the 15 score-population classes produced in step 2 aggregated to the scale of the aggregation input layer using the majority class.

RE Zone Raster Locations

This tab enables the user to obtain further insights from the results based on proximity to potential RE zones. The outputs from this tab are assigned the combine classification score to the RE zones input raster and extract the aggregated polygon/admin units that intersect with RE zones.

Combine Classification Input Layer: The combine classification output raster layer produced in step 2 of the "Enablement" tab.

Aggregated Combine Classification Input Layer: The aggregated combine classification output polygon layer produced in step 3 of the "Enablement" tab.

Potential RE Zones Input Layer: A raster layer highlighting the potetial RE zones. Pixels of zones in the region that have no RE potential needs to be represented with "no data" or "inf" values in the raster file.

Output (RE zones): A raster layer of the RE zones with the combine classification scores assigned to the highlited RE zones. Additionally, the aggregated polygon/admin units that intersect with RE zones are extracted along with their majority combine classification class to a new polygon layer.

RE Point Locations

This tab enables the user to obtain further insights from the results based on proximity to RE point locations. The first output from this tab aggregates the combine classification score based on the majority class that falls within the RE point location buffers. The second output extracts the aggregated polygon/admin units that intersect with RE point location buffers. The majority class extraction for the buffers is conducted using the original combine classification raster produced in step 2 of the "Enablement" tab.

Combine Classification Input Layer: The combine classification output raster layer produced in step 2 of the "Enablement" tab.

Aggregated Combine Classification Input Layer: The aggregated combine classification output polygon layer produced in step 3 of the "Enablement" tab.

RE Point Location Input Layer: Point locations of interest (.shp). This could be existing RE job locations or other points of interest.

Buffer Distance: Maximum radial distance of circular buffer from point location in meters.

Output (RE Point): A polygon layer showing the 15 score-population classes aggregated to the RE point buffers scale using the majority class that fall within the buffers. Additionally, the aggregated polygon/admin units that intersect with RE point location buffers are extracted along with their majority class to a new polygon layer.

Classes

Insights
The raw aggregate scores can be used in combination with information relating to the distribution of women and renewable energy sites to draw further insights. The insights tab groups population counts into three categories based on the lower, median, and upper quartile range of data to identify areas, of low, medium, and high numbers of women. These groupings are then combined with score classes to create 15 score-population classes listed below:
1 - Very low enablement, low population
2 - Very low enablement, medium population
3 - Very low enablement, high population
4 - Low enablement, low population
5 - Low enablement, medium population
6 - Low enablement, high population
7 - Moderately enabling, low population
8 - Moderately enabling, medium population
9 - Moderately enabling, high population
10 - Enabling, low population
11 - Enabling, medium population
12 - Enabling, high population
13 - Highly enabling, low population
14 - Highly enabling, medium population
15 - Highly enabling, high population

Updates 11 Dec 2024:

Colour palette for the 15 classes

Updated wireframe:

Subtickets for this issue:

• Generate 15 class WEE score given analysis aggregate and population as inputs #701
• Generate aggregate per subnational boundary by calculating the zonal stats mean of WEE #702
• Generate aggregate for point work opportunities layer #703
• Geenrate aggregate for polygon work opportunities layer #704
• Generate aggregate for raster work opportunities layer #705
• Implement UI for insights as per wireframes #706
• Implement unit tests for insights workflows #707

[mvmaltitz]

@timlinux Add a button next to the project button for Insights

[timlinux]

Alternative approach we could do where we implement it as a dedicated panel:

[timlinux]

Clipping by point (with buffer distances), polygon or raster