I have measured the spectral reflectance of a number of items. I wish to determine whether it is possible to identify these items by their reflectance alone. For example, by analysing spectra alone can I differentiate between a lemon and orange fruit?
To answer this question, I wish to conduct hierarchical clustering in R on my samples of spectra (to see if spectral measurements cluster by types of fruit, e.g,, an orange cluster, a lemon cluster, a banana cluster).
I have measured the spectral reflectance of a number of items. I wish to determine whether it is possible to identify these items by their reflectance alone. For example, by analysing spectra alone can I differentiate between a lemon and orange fruit?
To answer this question, I wish to conduct hierarchical clustering in R on my samples of spectra (to see if spectral measurements cluster by types of fruit, e.g, an orange cluster, a lemon cluster, a banana cluster).
How can I do this in R?
Note: please provide Tidyverse friendly suggestions as this is how I'm learning R and it would help me greatly.
Dataset
For each object, I have conducted several sample measurements using hand spectroradiometer. Here's an example DF with mock values. I want to see if the spectra for each object are different enough to identify the object.
Note: the real-life application would be to determine whether it's possible to determine these different fruit crops from satellite imagery using their spectral reflectance.
+------------+----------------+----------------+----------------+---------------+---------------+----------------+
| Wavelength | Orange_sample1 | Orange_sample2 | Orange_sample3 | Lemon_sample1 | Lemon_sample2 | Banana_sample1 |
+------------+----------------+----------------+----------------+---------------+---------------+----------------+
| 400 | 0.1 | 0.7 | 1.3 | 1.9 | 2.5 | 3.1 |
| 450 | 0.3 | 0.5 | 0.7 | 0.9 | 1.1 | 1.3 |
| 500 | 0.5 | 0.9 | 1.3 | 1.7 | 2.1 | 2.5 |
| 550 | 0.7 | 0.9 | 1.1 | 1.3 | 1.5 | 1.7 |
+------------+----------------+----------------+----------------+---------------+---------------+----------------+
How can I do this in R (preferably with a Tidyverse workflow)?
Dataset
For each object, I have conducted several sample measurements using hand spectroradiometer. Here's an example DF with mock values. I want to see if the spectra for each object are different enough to identify the object.
Note: the real-life application would be to determine whether it's possible to determine these different fruit crops from satellite imagery using their spectral reflectance.