Word Clouds for Different Varietals

Here is some more analysis, in visual form, of the soil and varietal analysis I did of last year's coffee crop in Puerto Rico. I entered each word every time it appeared on a cupping form, and correlated the data with different varietals. I have many of these charts made, but the files are large, so I'm posting just four here to give you an idea of how they look.

Most of these descriptors are aromatics, because the standard cupping forms have numerical scores for acidity, flavor, balance, and other major categories. Still, if a cupper (myself included) made a note of anything specific, no matter what the category, I included it in these diagrams.

The larger the word is, the more often it came up. The colors are included for contrast only — there is no data attached to the word colors. Click on the pictures to see them up close.

---

Fig. 7: descriptors for BOURBON Strong positive aromatics, including notable honey, chocolate, toast, and caramel. Strongest presence of the very positive sweet, as well as the positive tangy.

Fig. 4: descriptors for CATIMOR Dry is the dominant characteristic, indicated a dry, unpleasant mouthfeel or aftertaste. Ordinary nutty aromatics dominate, with malty tendencies. Negative peanutty and astringent notes are common.

Fig. 9: descriptors for TOP NINE COFFEES (all varietals) In all we find that the very best Puerto Rican coffees tend to be sweet, bright, tangy and smooth or delicate, with high incidence of honey, nutty, lemon, caramel, chocolate, toast, and floral aromatics

Fig. 10: descriptors for BOTTOM NINE COFFEES (all varietals) Overall, below-average to poor Puerto Rican coffees suffer from musty and peanutty aromatics, and show high incidence of sourness, dryness, and a lack of sweetness.

Comments

[Brandon Smyth]

This is awesome - nice work Daniel.

[Daniel Humphries]

Thanks Brandon. It was fun making them.

[Michael]

Daniel:

Once again, you have created something beautiful. This is really a pleasure to look at.

Lovely,

M

[Daniel Humphries]

Thanks Michael!

[Anunumous]

Daniel, the catimor coffees... I wonder how the samples differentiated in character (and score) from lots that were uniform in ripeness, to those that were not. Do you have any data, or even anecdata?

[Libby]

These are really cool! Thanks for sharing.

[Daniel Humphries]

Anun: We controlled for ripeness in this study. So all the varietals are at the same level of ripeness. We ended up making a Google Earth map of all this data. Next year my plan is to take digital photos of every sample as it's collected, before processing, showing cherry color. Then when you hover over a farm location on Google Earth, a picture will pop up showing cherry ripeness. Could also do this for the green coffee and even the roasted coffee. It's pretty easy to do. But the ripeness is what I really want to be shown in detail.

[Janet]

These are great visual representations of "flavor maps" for coffee. I have seen similar maps of chocolate flavors showing the subtle, and sometimes not so subtle, differences between cocoa varietals (I am a candy maker). I'm sure they exist for the wine appreciators out there, too, and whiskey drinkers.
My husband is a coffee grower here in Hawai'i and wonders how the roasting process influences the comments coming form the cuppers for this exercise. That is, were all the samples for these studies just taken to first-crack or were the profiles tailored to the varietal?
Aloha, Janet