Personal projects that feature a little bit of everything I've learned so far. Please check the Jupyter notebooks (.ipynb files) to see the full processes!
In this project I scrape all of the entries in a cat food review website, clean them and group them into a data frame, analyze them looking for patterns in its rating system, and design a model in order to predict cat food scores.
Skills:
- Web scraping
- Data cleaning
- Data analysis
- Training and evaluation of simple predictive models
Tools:
scrapy
pandas
matplotlib
andseaborn
scikit-learn
Here I try to implement myself some common machine learning models with the objective of understanding them properly.
Skills:
- math, statistics, machine learning theory in general
Tools:
numpy
scipy
matplotlib
andseaborn
scikit-learn
(mostly for comparison and for its convenient metrics functions)
A toy project where I import data from meteogalicia and look for recent tendencies in the levels of common air pollutants with the ojective of checking if the recent decrease in traffic (due to coronavirus) has impacted contamination significantly.
Skills:
- Request and processing of data accessed as JSON
- Data cleaning
- Data analysis
Tools:
requests
andjson
modulespandas
matplotlib
andseaborn
There are other projects that I've worked on that are either too small or simple to market them as "portfolio projects", or too big or specific to not have their own repository.
A command line application designed to generate and interpret the necessary files to perform a very specific chemical calculation (non planar Nucleus Independent Chemical Shift estimation, to be precise).
Skills:
- Parsing of long calculation outputs
- Working with 3D cartesian coordinates
- Fitting complicated 3D surfaces by combining simple 3D functions using a multiple linear regression model
Tools:
matplotlib
(includingmplot3d
)scikit-learn
- Design and development of a MySQL database to keep the inventory of a bakery business, as well as useful stored functions and transactions, and possibly some sort of UI to make it usable