Blog Posts by Andrew B. Collier / @datawookie

Cloudflare is a service that aims improve the performance and security of websites. It operates as a content delivery network (CDN) to ensure faster load times and consequently better user experience. However, it also protects against online threats by filtering “malicious” traffic.

Web scraping requests are often deemed to be malicious (certainly by Cloudflare!) and thus blocked. There are various approaches to circumventing this, most of which involve running a live browser instance. For some applications though, this is a bit hammer for a small nail. The cloudscraper package provides a lightweight option for dealing with Cloudflare and has an API similar to the requests package.

cURL is the ultimate Swiss Army Knife for interacting with network protocols. But to be honest, I really only scratch the surface of what’s possible. Usually my workflow is something like this:

1. Copy a cURL command from my browser’s Developer Tools.
2. Test out the cURL command in a terminal.
3. Convert the cURL command into a programming language (normally Python or R).
4. Prosper.

I’m going to take a look at my favourite online tool for converting a cURL command to code and then see what other tools there, focusing on Python and R as target languages.

The Big Book of R provides a comprehensive and ever-growing overview of a broad selection of R programming books. It was created and is maintained by Oscar Baruffa. The collection began with approximately 100 books and, with the help of contributions from the R community, has subsequently expanded to over 400. The books are grouped into topics such as geospatial, machine learning, statistics, text analysis, and many more. The Big Book of R is an excellent resource for anyone learning R programming, whether they are a beginner or advanced user.

The ability to specify a message ID in emails sent from the {emayili} package makes it possible to create email threads.

A new minor version of the openai-python package was released late on Friday 7 June 2024, only a couple of days after the last minor release. This release adds a chunking_strategy argument to the methods for adding files to vector stores.

Quick notes on installing Docker on various platforms:

• Linux
• macOS and
• Windows.

This question on Stack Overflow was a fun challenge: extract the markers off an embedded Google Map.

The R version of my Desert Island Docker talk. Similar idea to Desert Island Docker: Python Edition.

Over the years that I’ve been dabbling in public speaking I’ve generally developed a talk, presented it once and then moved on. However, I’ve noticed other speakers who give the same (or similar) talk at different events, where the talk evolves and improves over time.

From time to time I want to extract the table of contents from a PDF. Here’s how I do that using simple shell tools.

The {emayili} package has adapters which make it simple to send email via a variety of services. For example, it caters specifically for ZeptoMail, MailerSend, Mailfence and Sendinblue. The latest version of {emayili}, 0.8.0 published on 23 April 2024, adds an adapter for Mailtrap.

The key to successful backtesting is to ensure that you only use the data that were available at the time of the prediction. No “future” data can be included in the model training set, otherwise the model will suffer from look-ahead bias (having unrealistic access to future data).

Are you trying to build and push a Docker image from GitHub Actions?

The Two-Fund Separation Theorem introduced by James Tobin, a Nobel Prize-winning economist, is a fundamental concept in investment theory. It addresses how investors can optimally allocate their assets. In an efficient market an optimal portfolio is a combination of a risk-free asset and a market portfolio.

Sprinkling status messages across you code using print() statements can be a good temporary fix for tracking down issues in your code.

But it can get messy and there’s no way for you to selectively disable them. Sure you can redirect output to a file or /dev/null but this is an all-or-nothing solution. How about disabling some messages and retaining others?

This is where the logging module comes into its own.

Setting the values of one or more parameters for a GARCH model or applying constraints to the range of permissible values can be useful.

The two quantities we have been modelling (the time-dependent average and standard deviation of the returns) represent respectively the (potential) risk and reward associated with an asset. The relationship between these two quantities is implicit in the GARCH model. However, sometimes the return depends directly on the risk. A variant of the GARCH model can take this explicit relationship into account.

Most often when you are creating a new Docker image it will be based on one of the standard Docker base images like ubuntu, alpine, python or nginx. But sometimes you might want to truly roll your own image. Starting with literally nothing. From scratch. Tabula rasa.

Is this a “good” model? How to validate a model and determine whether it’s a good representation of the training data and likely to produce robust and reliable predictions.

In general a parsimonious model is a good model. A model with too many parameters is likely to overfit the data. How do we determine when a model is “complex enough” but not “too complex”?