1 front-end developer:
problems
mockups to websites
fast loading speeds
different kinds
Application focus: JavaScript
essential skills
Breaking problems into manageable chunks.
frameworks
2 back-end programmer:
The technologies used by back-end programmers tend to stick around longer than those used on the front-end because rearchitecting an application is a major undertaking. As a result back-end technologies change more slowly.
Languages like Java and Go are popular among large teams and large projects because they enforce code organization and structure. Languages like Python, Ruby and Javascript allow rapid development and easy prototyping, but they don’t necessarily scale to heavy loads as easily. This implies cultural differences between the communities around different languages, and in the teams that use the different languages. Java in particular is used at larger enterprises. PHP tends to be seen in legacy applications.
3 Mobile programmer: iOS, Android
problems
clean user interface
sync to server
from a functional state to a production ready state
mobile hardware
essential skills
programming paradigm called model view controller
ios: swift (since 2014 summer), objective C, IDE is Xcode
android: Java, IDE is Android Studio
4 data analyst.
problems
applied statistics, machine learning
data visualization
In her view, data analysts is to basically help other people in the company make decisions and prioritize their work by using the data we collect
language: sequel to fetch data, or similar language (Hadoop or MapReduce), statistical package like R, python, whatever for more advanced statistics and modeling and numerical analysis
essential skills
- Mathematics and statistics. Understand how to interpret the results, the technical aspects of models that are employed, develop new or alternate techniques to go beyond the built-in.
- Programming. Beyond software package to perform “out of box”. customized effects
- a sense of curiosity. Ask questions of their data to generate a logical flow in analysis. If an oddity is detected, such as missing data, outliers, unexpected trends, steps should be taken to understand the oddity and try to resolve it.
python vs R
https://www.dataquest.io/blog/python-vs-r/
R. Best at very specialized task:
- ggplot2: good-looking visualizations to hasten exploration of the data
- dplyr, tidyr: reshaping data
Python. easier to learn and understand
- scikit-learn, matplotlib, seaborn: for machine learning and visualization
learning zone:
The Safe Zone - This is where concepts are familiar and we don’t have to take any risks. The Safe Zone is important because it gives us a place to return and reflect on our current understanding. However, our potential for growth is limited in the Safe Zone because we are not challenged with unfamiliar concepts.