It’s no secret that working with tech can be incredibly frustrating.
While computers are able to process instructions in a matter of seconds, actually getting to that point can sometimes take hours, or even days. When you’re trying to work with code and you hit a snag it’s tempting to blame the computer but the truth is that computers only do what their told to do so it’s likely that the problem is something that you did, or didn’t, do. Inability to get something to work, like a piece of software, or a line of code is one of the biggest reasons people quit learning to code or work with machine learning algorithms.
Having spent many, many evenings searching for solutions to my tech problems, I’ve created this page as a repository of problems that I’ve encountered and how I (eventually) managed to fix them. Hopefully it is useful to anyone out there struggling with some sort of IT problem! You might want to bookmark this page as I update it, as and when, I come across new problems and their solutions.
Whenever you have a problem with your computer, and it isn’t immediately obvious why, there are a handful of steps that you should take that may (hopefully!) clear the problem up quickly.
Plug your computer in
If you’re working on a laptop that isn’t plugged in and:
- It’s really slow
- Lagging between tasks
- Generally behaving a little buggy
Go ahead and plug it in. This is the simplest of fixes but it might be that your laptop battery isn’t providing as much power as your computer requires to execute processes smoothly.
Clear your browser cache
For web-based work, a common problem you might encounter is a stale browser cache. If you ever contact a help-desk this is almost certainly going to be their first suggestion.
When you visit a web page, the contents of that page are downloaded and stored locally on your computer in what’s known as a “cache”. This means that when you visit the website again the browser doesn’t need to re-download the information, thereby speeding up the amount of time it takes for your browser to load the page. If you make a change to a website the cached information might be out-of-date.
If you’ve made a change and don’t see it reflected on the website, try refreshing your cache. The first thing to try is
CTRL + R (PC)
CMD + R (Mac), which refreshes your browser. In some instances this is enough. If you’re still encountering problems try a hard refresh. On a PC, try
CTRL + F5. On a Mac, instructions depend on which browser you’re using, for Chrome or Firefox try
⌘ CMD + ⇧ SHIFT + R, for Safari try
⇧ Shift and the Reload toolbar button. If none of the above fix the problem then you can clear your browser’s cache manually. Instruction on how to do this can be found here for Chome, Firefox, or Safari.
Try another browser
If you’re working in a browser and it’s behaving strangely and clearing the cache hasn’t helped, try another browser. So if you’re using Chrome try Firefox and if you’re using Safari or Edge try Chrome or Firefox. Sometimes there are conflicts between what you’re trying to do and the code of the browser you’re using.
Administrator mode is more powerful than the standard mode that your computer runs in as it provides you with full access to your computer, allowing you to add, or delete, files and programs. If you are, for example, trying to
pip install a Python package and continually get an error message try running the command prompt in administrator mode. It’s a simple tweak that could save you hours of very frustrating troubleshooting.
On Windows this is done by right-clicking on the program that you want to run and selecting “Run as administrator”. Mac doesn’t have an administrator mode in the same was Windows, however, if you’re using Terminal you can use the
sudo command to run commands as an administrator. Your command would look something like this:
$ sudo some command here
Uninstall and Reinstall
If you’ve tried all of the above steps and are still having problems with a piece of software or app, then you might want to try uninstalling and then reinstalling it. It’s possible that the way you’ve configured the software or a recent update is causing conflicts with other things on your computer. The downsides of this approach are that you will, of course, lose any configurations or changes made inside of the software. Make sure you back up any related data before uninstalling.
Troubleshooting Specific Problems
General troubleshooting tips are great when they work but often the problem lies elsewhere. Below are some of the specific problems that I’ve encountered and how I fixed them.
I use Jupyter Notebooks alot, for both work and my own research and it’s a great way to write, execute, and annotate code. Recently, however, I was working on a notebook with a deadline when my Jupyter workflow ground to a complete halt.
The notebook wasn’t executing my code. Nothing was happening.
Sometimes running a simple line of code would take minutes, sometimes it wouldn’t run at all. I could tell that the notebook was unresponsive because the command prompt was not showing any activity. After many, many hours of troubleshooting I eventually found that my code would run if I first turned the wifi off. I thought this was strange since the Jupyter server runs locally but I had a deadline!
After a few hours of turning the wifi off to run my code and then back on when I needed to search for something online, the notebook that I had been working on became corrupted and I was not able to access either my file or the checkpoint backup.
Problem: Jupyter was not executing code, it was unresponsive and occasionally crashing. There was no error message. It then entirely corrupted a file.
Solution: My Jupyter Notebook application was installed through Anaconda. Generally, Anaconda is recommended because it comes pre-installed with the most common machine learning packages and is an easy way to get started quickly. In this instance, though, the Anaconda distribution of Jupyter was causing some conflicts which was making it impossible for me to complete my task. I uninstalled Anaconda’s version of Jupyter and reinstalled it directly to my machine using `pip`. I was able to gain access to the checkpoint backup of my file once I had Jupyter up and running again, although the original still remains unusable.