Repository navigation
Conversation
JarettBakerDunn
left a comment
JarettBakerDunn
left a comment
Collaborator
There was a problem hiding this comment.
I just found a few small typos. otherwise this looks good! everything ran and plotted fine. Although if we want the plotted images to appear in the documentation we will need to create cells which just display the prerendered image.
| "source": [ | ||
| "### 3. Preprocessing: standardization\n", | ||
| "\n", | ||
| "As we discussed above, we center each selected oxide feature so is has zero mean so that PCA does not simply point toward the mean of the data. Additionally, we divide each feature by its standard deviation so that the features are scaled to unit variance. This is important because oxide features are measured in different ranges (e.g., SIO2 is often 40-70 wt% while TIO2 is often <1 wt%). If we do not scale the features, PCA will be dominated by the variation of SIO2.\n", |
| "source": [ | ||
| " <div style=\"text-align: right\"> \n", | ||
| " → <b>NEXT: [t-SNE on geochemical data](./2_manifold_learning.ipynb) </b> <a href=\"\"></a> \n", | ||
| " <img src=\"../assets/education-gem-notebooks_icon.png\" alt=\"icon\" style=\"width:4%\">\n", |
Collaborator
There was a problem hiding this comment.
the icon is not appearing for me, assets folder is three levels above here, so I think this should be "../../../assets/education-gem-notebooks_icon.png"
| "\n", | ||
| "Here:\n", | ||
| " - V is the matrix of principal-component directions (eigenvectors) of size $\\{p \\times m\\}$, where $m$ is the number of principal components we want to keep, usually $m < p$ since we want to reduce the dimensionality of the data.\n", | ||
| " - Z is the matrix represents the data projected onto the principal components and is of size $\\{n \\times m\\}$.\n", |
Collaborator
There was a problem hiding this comment.
"Z is the matrix represents" -> "Z is the matrix that represents..." ?
| "\n", | ||
| "To avoid this, we retain only rows with measurements for all selected features before fitting PCA. This complete-case choice would reduce the dataset size.\n", | ||
| "\n", | ||
| "We keep `ROCK TYPE` and `TECTONIC SETTING`for visualizing the PCA scores. These describe the database's geological context and should not be treated as labels validated independently of the geochemical measurements." |
Collaborator
There was a problem hiding this comment.
space between TECTONIC SETTING and for ?
| "source": [ | ||
| "### 5. Understanding PC scores by tectonic setting\n", | ||
| "\n", | ||
| "Each point in the PC1-PC2 score plot represents one mineral analysis, colored by its recorded tectonic setting. The plotted setting is directly from the database, not a prediction by PCA. Through this visualization, we can explore samples from different recorded geological contexts form clusters or occupy different regions of score space." |
Collaborator
There was a problem hiding this comment.
should this be "we can explore how samples..."?
| "\n", | ||
| "- **HIMU** : mantle source with a high time-integrated U/Pb ratio ($\\mu = {}^{238}\\text{U}/{}^{204}\\text{Pb}$), characterized by very radiogenic Pb (high $^{206}\\text{Pb}/^{204}\\text{Pb}$), low $^{87}\\text{Sr}/^{86}\\text{Sr}$, among other element ratios, observed in some OIBs (e.g., St. Helena, Cook-Austral islands),\n", | ||
| "\n", | ||
| "- **EM1** : Enriched Mantle 1, with lower $\\text{U}/\\text{Pb}$ and lower $^{206}\\text{Pb}/^{204}\\text{Pb}$ than HIMU, intermediate $^{87}\\text{Sr}/^{86}\\text{Sr}$ values, distinct from EM1 and often attributed to recycled ancient oceanic crust materials observed in some OIBs (e.g., Kerguelen, Tristan),\n", |
Collaborator
There was a problem hiding this comment.
distinct from EM1 -> distinct from EM2?
| }, | ||
| "outputs": [], | ||
| "source": [ | ||
| "# The starting poin\n", |
Collaborator
There was a problem hiding this comment.
starting poin -> starting point?
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR adds the notebook on linear and nonlinear dimensionality using pca and t-sne methods on the geochemical data. The current state addresses the suggestions made on the internal google drive.