The present projects have been developped in the course of following Udacity's NanoDegree in Data Analysis.
During the year 2019, around 100k medical records have been registered and made public on Kaggle. The doctors hoped analyses would reveal patterns that would permit them to predict, to a certain precision, when a patient would be likely to show-up to their appointment. My analysis aimed at identifying features that would serve that purpose. It was carried out by trying to answer the following quesitons :
- Did the absentees receive messages prior to their appointments ?
- Did the difference between the appointments' dates and their scheduling contribute to the absences ?
- What are the most common medical conditions among the absentees ?
- Is there an age range with a high absence rate ?
- Among which gender is absence more frequent ?
- Between the beneficiaries and non-beneficiaries of the Bolsa Familia program, which category was absent the most ?
The archived tweets of @WeRateDogs are collected from different sources (flat file, via an url and twitter's API) and put together. The main purpose is assessing and cleaning the datasets, preparing them for an analysis.
The latter involved finding out the 10 most appreciated dogs and their species. FInally, the results of the analysis are presented as a blog article addressed to dog lovers.
This is an exploratory and explanatory data analysis of the Ford Gobike dataset. The latter comprised 183 412 data on bike trips that took place in San Francisco's Bay Area in February 2019 via the Ford Gobike bake sharing system. It starts with some necessary data wrangling and feature engineering as see fit to carry out EDA, which covered univariate, bivariate and multivariate analysis.