GRAPHICAL REPRESENTATION AND TWO GROUPS ANALYSIS ON DATA MATRIX OF ROBUSTA GREEN CHERRIES PRODUCTION IN TWO HARVEST PERIODS

Several factors that play a role in the productivity of Robusta coffee trees are the influence of pruning techniques and weather elements. This paper discussed the graphical analysis and comparison of two data matrices of Robusta green cherries production, which would enter the ripening process in branch categories for the harvest period in 2023 and 2024. Hypothesis testing on secondary data in the form of daily weather conditions in 2022 and 2023, which include temperature, dew, humidity, wind speed, and cloud cover for the two periods, was significantly different. However, solar radiation and precipitation were not. The data source for each harvest period was primary data, with the object being a sample of 30 trees that were sampled purposively. The research object was in Pagaralam Municipality, South Sumatra. There were 18 variables covering many branch categories based on production year, position, and shape. The PCA (Principal Component Analysis) results on each data matrix show similarities in the dominant variables representing each subspace. The first three PCs in each data matrix for 2023 and 2024 span a subspace and describe the variation of the original data of 77.3% and 68.8%, respectively. The 3rd and 1st-year production branch categories dominate the subspace of each data matrix for 2023 and 2024. Comparison of the two PC subspaces using two groups analysis in 3rd dimension space produces angles of 19.70, 28.80, and 69.10. The bisector components show that the variables that dominate the similarity of the two data matrices are the variables that tend to represent both PC subspaces dominantly. Robusta green cherry production can be represented by the number of secondary branches, which are straight in shape, along with the number of fruit clusters. This study result can be a reference for farmers when considering the composition of the number of branch categories when pruning.