Note: you opted to see also "groups and other broadly defined products" (left panel checkbox). This means that losses for groups (e.g., "Cereals") and other generic commoditiy groups (e.g., "Other citrus fruits, n.e.c.") are also shown, which may bias the statistics by stage. Thus, either remove that option, or be careful in interpreting the results below.
Potential duplicates
By default a duplicate is defined as an entry with the same M49 code, CPC code, stage, year, fsc_location (i.e., stage), and loss factor. You can define youw own list of variables (e.g., if the same value is shared by different regions within the country you should need to add also the region to the comparison so that these get excluded).
The first tab shows the number of entries for each combination of selected variables, which may indicate a duplicate. The second table shows the actual entries, grouped by potential duplicate (groups are separated by yellow delimiters).
Summary
Entries that can be potential duplicates
Extreme values (or outliers)
Loss percentages that are relatively small or big are reported below. The thresholds that define a "small" or "big" percentage are defined by the "Percentile" control below. Percentiles are calculated on observations on all countries and crops, the idea being spotting entries that are extreme at the global level. You can choose to remove thresholds that refer to a very small subset of observations by selecting the minimal amount of observation witht the "Remove combinations ..." control.
Refresh dataset
Clicking on the button below will stop the app and refresh the dataset.