Please share with the community what you think needs improvement with Anaconda.
What are its weaknesses? What would you like to see changed in a future version?
The solution's support is important and needs to be better. I don't have the last update due to the fact that when I tried to update it I had an error and ran into issues. It's not just me; lots of people in the community don't have the last update. If support was better they may be able to address issues like this faster. The stability could be improved. Stability is very important because if you develop some product or some program, you want a very, very stable software that you can use for more than two or three years. When you install Anaconda for the first time, it's really difficult to update it. I can't think of any features that are lacking. Overall, it works quite well for me.
It crashes once in a while. In case of a reboot or something unexpected, the unseen code part will get diminished, and it relatively takes longer than other applications when a reboot is happening. They can improve in these areas. They can also bring some database software. They have software for analytics and virtualization. However, they don't have any software for the database.
The product can be improved in a few ways. It would be possible to simplify the installation although it was not a problem in my case because of my experience. One thing that hurts the product is that the company is not doing more to advertise it as a solution and make it more well known that I have seen. I do not really feel it is as known as it should be in our market. The features I would like to see in the next release are more packages. That is, it would be nice to have more libraries added by default.
Anaconda has a platform for data science professionals, but if they would have something for an on-prem solution for the data science platform, it would be more valuable to Anaconda. The user interface could be improved. There should be easy provisioning on AWS. Configuring clusters could be easier. The solution would benefit from offering more automation.
Having a small guide or video on the tool would help learn how to use it and what the features are.
The interface could be improved. Other solutions, like Visual Studio, have much better UI. In Visual Studio, we can create projects. I'm not sure it's that kind of feature is possible on Anaconda. In Visual Studio, we can bring a solution and then we can keep multiple projects organized and into it. If there are any libraries, we can have them as well and we can just call them. We can easily link the libraries into the project. This should be possible on Anaconda as well.
I think that the framework can be improved to make it easier for people to discover and use things on their own. They need a better interface because currently, we have to do everything through coding. It would be nice to have a simple description of what each library is used for and how to use it. I would like to see additional libraries included to support computer vision and natural language processing. The framework gives us the ability to create them, but having more in place would mean that we would need to do less coding.
I would like to see the inclusion of some statistical modeling functionality. Having some examples built-in that we can customize based on the use case, rather than having to build the entire model, would really be an advantage. Additional support for the visualizations would be an improvement. The ability to schedule scripts for the building and monitoring of jobs would be an advantage for this platform.
One feature that I would like to see is being able to use a different language in a different cell, which would allow me to mix R and Python together. For example, using R tidyverse to wrangle data, R ggplot to visualise data and then use Python sci kit-learn to build machine leanring model on that data. Multi-language support would allow all of the data science languages to be in one place, and we could become a hub for it.
I have nothing to say about improvements; I love this product. It's our bread and butter and we use it every day. We did have issues with virtual environments in the past, but that has worked itself out.
I hit some contribution issues and attribution problems, so the product could be improved in that area. There's always room for improvement. It's one thing if you have an IT guy with the solution, but there were some cases when it wasn't so simple. There were a lot of typos in the documentation and because we were using the product on-premise, the solution had to be implemented by the IT team here and they had some difficulty fixing problems, particularly from the wall sheet. I think better documentation or a step-by-step guide for installation would help, especially for on-premise users. That would be great. I haven't used it enough to think about additional features and I didn't hit any roadblocks that made me think about that. It worked well for me.
* Currently, it's working perfectly on Microsoft Windows but lags in open source OS like Ubuntu, Mint. * File uploading feature needs to be improved.
When setting up an ML platform, can either Anaconda or SageMaker be used in an AWS deep learning machine for ML?
Is one better than the other?
Let the community know what you think. Share your opinions now!