This tool provides users with a solution to select relevant areas and locations where they can conduct the validation trials for their Minimum Viable Products (MVP’s). The tool guides the user through a set of steps that will allow them to make decisions on what is the most relevant set of environmental clusters that can help reduce variance due to environmental conditions, making the results of their trials more representative for their target area (AOI).
On the left-side panel of the tool, the user is initially provided with a set of options to define the trial conditions in which they want to run validations.
The user will need to provide the file in .shp or .geojson formats to set the AOI where the trials will be conducted. This will define the tool for which to process the different environmental units in the AOI.
Alternatively, the user could manually draw a shape on the map that defines the region of interest.
Note: The size of the area has a limit of XXX km2.Consider this, as the tool will not work for very large areas. In case you need to run the tool for a very large AOI, you can contact xxxx.xxxxx@cgiar.org
In this step, the user will define the main crop that the trials are targeting to validate. This is done to select a set of relevant environmental variables that are considered most relevant by agronomists. The crops available, together with the list of associated variables are indicated in the table below:
[Table with crops + variables]
Seasonal conditions affect how environments might be clustered within the AOI, thus affecting the results of the trials and how the crop performs. The tool requires that the user indicates the seasonality (start month and end month) when the trials will be conducted.
After defining the trial conditions, the user will be provided with a first result of the optimal number of clusters in the AOI. These indicates which are the environmentally distinct areas specific for the crop the trials are targeting.
[To be added… The user can view this clusters and also see what is the proportion of trials that should be set-up in each cluster to get a balanced sample size that improves representativeness of the results.]
[Below the map the user can see the boxplots for the different variables included for the crops and which are the distributions (in boxplots) of the variables within each cluster.]
If the user is satisfied with the results, these can be exported to further investigation or discussion with the team. Click the “Get Results” button to download to your computer.
In case the user does not agree with the results, is possible to navigate to the next step, where the user will be able to create a more tailored set of clusters. This optional step strongly influences the results.
If for some reason the optimal number of clusters defined by the dry-run of the tool is not desirable, the user may indicate what would be a more adequate number that fits their knowledge of the area or the logistical constraints. Please, [select a number from the drop-down / indicate a number].
At this point, the user may also decide which are the most relevant environmental variables that may influence the trials or that are of specific focus in the validation. For example, an MVP might be focusing on drought-resistant varieties in rainfed wheat, and therefore rainfall might be the most important environmental variable to define the different environmental clusters.
The user should remove the variables from the box and only leave those that are relevant for the validation.
The user can test several combinations of variables and number of clusters and see the effects in the clusters and variable distributions until an acceptable result is obtained. Once satisfied, the user can export the user-defined clusters by clicking the “Get Results” button to download to the computer.