From f54de2ebf85724b1d742fd20b392c62b016dd1f7 Mon Sep 17 00:00:00 2001 From: WeCo bot <53337611+weco-bot@users.noreply.github.com> Date: Tue, 17 Sep 2024 09:02:13 +0000 Subject: [PATCH] Update notebooks --- docs/examples/01-exploring-wellcome-collections-apis.md | 2 +- docs/examples/02-extracting-more-data-for-local-analysis.md | 2 +- docs/examples/03-connecting-the-apis-together.md | 2 +- docs/examples/04-building-graphs-of-visually-similar-images.md | 2 +- docs/examples/05-working-with-snapshots-of-the-api.md | 2 +- docs/examples/06-visualising-the-collection-on-a-map.md | 2 +- docs/examples/07-building-an-image-classifier.md | 2 +- docs/examples/08-extracting-features-from-text.md | 2 +- docs/examples/09-building-a-recommender-system-for-subjects.md | 2 +- 9 files changed, 9 insertions(+), 9 deletions(-) diff --git a/docs/examples/01-exploring-wellcome-collections-apis.md b/docs/examples/01-exploring-wellcome-collections-apis.md index 596f891..e208f49 100644 --- a/docs/examples/01-exploring-wellcome-collections-apis.md +++ b/docs/examples/01-exploring-wellcome-collections-apis.md @@ -1,6 +1,6 @@ # 1. Exploring Wellcome Collection's APIs -[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/01-exploring-wellcome-collections-apis.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/01-exploring-wellcome-collections-apis.ipynb) +[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/01-exploring-wellcome-collections-apis.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/01-exploring-wellcome-collections-apis.ipynb) Wellcome collection has a few public APIs which can be used to fetch things like works, images, and concepts. They all live behind the following base URL diff --git a/docs/examples/02-extracting-more-data-for-local-analysis.md b/docs/examples/02-extracting-more-data-for-local-analysis.md index 3cffb01..6806d96 100644 --- a/docs/examples/02-extracting-more-data-for-local-analysis.md +++ b/docs/examples/02-extracting-more-data-for-local-analysis.md @@ -1,6 +1,6 @@ # 2. Extracting more data for local analysis -[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/02-extracting-more-data-for-local-analysis.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/02-extracting-more-data-for-local-analysis.ipynb) +[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/02-extracting-more-data-for-local-analysis.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/02-extracting-more-data-for-local-analysis.ipynb) In the last notebook, we saw that the `/works` API can do some clever querying and filtering. However, we often have questions which can't be answered by the API by itself. In those cases, it's useful to collect a load of data from the API and then analyse it locally. diff --git a/docs/examples/03-connecting-the-apis-together.md b/docs/examples/03-connecting-the-apis-together.md index 24f592b..8e12e90 100644 --- a/docs/examples/03-connecting-the-apis-together.md +++ b/docs/examples/03-connecting-the-apis-together.md @@ -1,6 +1,6 @@ # 3. Connecting the APIs together -[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/03-connecting-the-apis-together.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/03-connecting-the-apis-together.ipynb) +[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/03-connecting-the-apis-together.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/03-connecting-the-apis-together.ipynb) So far, we've only looked at the `/works` API, but Wellcome Collection has a few more which we can make use of. As well as `/works`, we can also use `/images` and `/concepts`. diff --git a/docs/examples/04-building-graphs-of-visually-similar-images.md b/docs/examples/04-building-graphs-of-visually-similar-images.md index 4a215ec..fb443df 100644 --- a/docs/examples/04-building-graphs-of-visually-similar-images.md +++ b/docs/examples/04-building-graphs-of-visually-similar-images.md @@ -1,6 +1,6 @@ # 4. Building graphs of visually similar images -[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/04-building-graphs-of-visually-similar-images.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/04-building-graphs-of-visually-similar-images.ipynb) +[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/04-building-graphs-of-visually-similar-images.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/04-building-graphs-of-visually-similar-images.ipynb) In the last notebook, we introduced the ability to fetch visually similar images using the `/images` API. diff --git a/docs/examples/05-working-with-snapshots-of-the-api.md b/docs/examples/05-working-with-snapshots-of-the-api.md index 4b63dfc..73dbcf0 100644 --- a/docs/examples/05-working-with-snapshots-of-the-api.md +++ b/docs/examples/05-working-with-snapshots-of-the-api.md @@ -1,6 +1,6 @@ # 5. Working with snapshots of the API -[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/05-working-with-snapshots-of-the-api.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/05-working-with-snapshots-of-the-api.ipynb) +[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/05-working-with-snapshots-of-the-api.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/05-working-with-snapshots-of-the-api.ipynb) As we saw at the end of the last notebook, the API limits its responses to 10,000 total results - after that point, users are directed to work with the snapshots. For example, making a request to gives us: diff --git a/docs/examples/06-visualising-the-collection-on-a-map.md b/docs/examples/06-visualising-the-collection-on-a-map.md index 37c1062..8cb2ce4 100644 --- a/docs/examples/06-visualising-the-collection-on-a-map.md +++ b/docs/examples/06-visualising-the-collection-on-a-map.md @@ -1,6 +1,6 @@ # 6. Visualising the collections on a map -[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/06-visualising-the-collection-on-a-map.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/06-visualising-the-collection-on-a-map.ipynb) +[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/06-visualising-the-collection-on-a-map.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/06-visualising-the-collection-on-a-map.ipynb) In this notebook, we're going to use a secondary API to visualise the geographical extent of the collection on a map. diff --git a/docs/examples/07-building-an-image-classifier.md b/docs/examples/07-building-an-image-classifier.md index 94b1f14..d8881df 100644 --- a/docs/examples/07-building-an-image-classifier.md +++ b/docs/examples/07-building-an-image-classifier.md @@ -1,6 +1,6 @@ # 7. Building an image classifier -[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/07-building-an-image-classifier.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/07-building-an-image-classifier.ipynb) +[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/07-building-an-image-classifier.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/07-building-an-image-classifier.ipynb) This notebook is going to race through some high-level concepts in machine learning (specifically, fine-tuning a convolutional neural network). However, our focus will remain on demonstrating the practical uses of the Wellcome Collection API. As such, some important ML topics will be covered in less detail than they deserve, and some will be skipped entirely. If you're not already familiar with the basics of ML but want to learn more, I'd recommend exploring [Practical Deep Learning for Coders](https://course.fast.ai/) by fast.ai. It describes itself as: diff --git a/docs/examples/08-extracting-features-from-text.md b/docs/examples/08-extracting-features-from-text.md index b4f00d7..6d2b93c 100644 --- a/docs/examples/08-extracting-features-from-text.md +++ b/docs/examples/08-extracting-features-from-text.md @@ -1,6 +1,6 @@ # 8. Extracting features from text -[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/08-extracting-features-from-text.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/08-extracting-features-from-text.ipynb) +[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/08-extracting-features-from-text.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/08-extracting-features-from-text.ipynb) In the last notebook, we saw that using a pre-trained network allowed us to extract features from images, and train a classifier for new categories on top of those features. We can do the same thing with text, using a pre-trained network to extract features from text. In this notebook, we'll use those features the visualise the similarities and differences between works in the collection, and try to find clusters of related material. diff --git a/docs/examples/09-building-a-recommender-system-for-subjects.md b/docs/examples/09-building-a-recommender-system-for-subjects.md index 3852417..4228538 100644 --- a/docs/examples/09-building-a-recommender-system-for-subjects.md +++ b/docs/examples/09-building-a-recommender-system-for-subjects.md @@ -1,6 +1,6 @@ # 9. Building a recommender system for subjects -[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/09-building-a-recommender-system-for-subjects.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/tree/rk/gha-build/notebooks/09-building-a-recommender-system-for-subjects.ipynb) +[View on GitHub](https://github.com/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/09-building-a-recommender-system-for-subjects.ipynb) | [Run in Google Colab](https://colab.research.google.com/github/wellcomecollection/developers.wellcomecollection.org/blob/rk/gha-build/notebooks/09-building-a-recommender-system-for-subjects.ipynb) Finally, we'll consider building a recommender system using data from Wellcome Collection. Thes machine learning models work slightly differently to the ones we've seen so far. Rather than being trained to predict a single value, they're trained to predict a whole matrix of interactions between two kinds of entity.