Difference between revisions of "Recipe/Example/Ratatouille cuisinée"
Jump to navigation
Jump to search
(7 intermediate revisions by the same user not shown) | |||
Line 27: | Line 27: | ||
sel - Salt, white (sea, igneous or rock), no enrichment (11017) (0,0,0,0,0,0,97.8) | sel - Salt, white (sea, igneous or rock), no enrichment (11017) (0,0,0,0,0,0,97.8) | ||
== Results == | == Results == | ||
+ | === NNLS === | ||
+ | Ratatouille recipe fit using nnls with forced order. | ||
Deviance: 0.11 (very good) | Deviance: 0.11 (very good) | ||
<gallery> | <gallery> | ||
− | + | RatatouilleRecipeFit.png | Blame | |
− | + | RatatouilleRecipe.png | Recipe | |
+ | RatatouileRecipePredictedVersusProduct.png | Recipe check | ||
</gallery> | </gallery> | ||
+ | The recipe check plot shows that the fit cannot predict the correct fractions for all vegetables. The nutritional values for these are too similar. Only because we add olive oil, most vegetables are forced to a fraction. But this fraction is not larger than olive oil. All the rest stays with the eggplant. | ||
+ | ==== Ingredient families ==== | ||
+ | * Eggplant/Zucchini - these have the same fraction. Zucchini forces Eggplant to the same fraction. Correlation is 0.88. Could the eggplant be higher? Can Zucchini be lower and will it be replaced by another vegetable? | ||
+ | * Tomato puree/Onion/Sweet peppers/Tomato concentrate - they are forced by Olive oil. These ingredients are highly correlated. Could their fractions be higher? | ||
+ | |||
+ | === John fit comments === | ||
+ | This contains a lot of vegetables, which will have similar nutrient profiles, so there is the potential for a wide variance. Including carbohydrates in the analysis gives some improvement. |
Latest revision as of 09:54, 13 February 2025
Product
Product Nutrition:
(5.4,0.7,5.8,5.8,2.1,1.5,0.62)
Ingredients:
Dés d'aubergines préfrites 25% (aubergines 22%, huile de tournesol), dés de courgettes 25%, purée de tomates mi-réduite 20%, oignons en dés 13%, poivrons rouges en dés 8%, double concentré de tomate 3%, huile d'olive vierge extra 2%, sucre 1,3%, sel, ail, jus concentré de citron, farine de BLÉ, thym 0,1%.
Ciqual
Aubergines - Eggplant, raw (20053) (0.14, 0.034, 2.39, 2.39, 2.7, 1.12, 0)
Courgettes - Courgette or zucchini, pulp and peel, raw (20020) (0.26, 0.061, 1.8, 1.79, 1.05, 1.23, 0.023)
Purée de tomates mi-réduite - Tomato puree, canned (20170) (0.4, 0.085, 7.54, 3.64, 2.38, 1.4, 0.51)
oignons en dés - Onion, raw (20034) (0.62,0.2, 6.25, 4.8, 1.7, 1.1, 0.098)
poivrons rouges en dés - Sweet pepper, green, yellow or red, raw (20041) (0.27, 0.071, 4.55, 2.7, 1.5, 0.8, 0.021)
double concentré de tomate - Tomato paste, double concentrate, canned (20268) (0.29, 0, 17, 16.5, 3.62, 3.73, 0.93)
huile d'olive vierge extra - Olive oil, extra virgin (17270) (99.9, 15.2, 0,0,0,0,0)
sucre - Sugar, white (31016) (0,0,99.8,99.8,0,0, 0.0054)
sel - Salt, white (sea, igneous or rock), no enrichment (11017) (0,0,0,0,0,0,97.8)
Results
NNLS
Ratatouille recipe fit using nnls with forced order. Deviance: 0.11 (very good)
The recipe check plot shows that the fit cannot predict the correct fractions for all vegetables. The nutritional values for these are too similar. Only because we add olive oil, most vegetables are forced to a fraction. But this fraction is not larger than olive oil. All the rest stays with the eggplant.
Ingredient families
- Eggplant/Zucchini - these have the same fraction. Zucchini forces Eggplant to the same fraction. Correlation is 0.88. Could the eggplant be higher? Can Zucchini be lower and will it be replaced by another vegetable?
- Tomato puree/Onion/Sweet peppers/Tomato concentrate - they are forced by Olive oil. These ingredients are highly correlated. Could their fractions be higher?
John fit comments
This contains a lot of vegetables, which will have similar nutrient profiles, so there is the potential for a wide variance. Including carbohydrates in the analysis gives some improvement.