A brief introduction of the word prediction app

Exploratory analysis of three textual datasets that will be used to create a simple word prediction algorithm.

Reproducible pitch on the "Guess the regression line!" Shiny Web Application.

n this analysis, data from U.S. National Oceanic and Atmospheric Administration’s (NOAA) is used to investigate the impact of weather events on both public health and economic consequences. The data is cleaned up by resolving the exponents of both property and crop damage to numeric values. These are then used together with the corresponding mantissa values to calculate the damage on property and crop. Then, the labeling of the weather events is cleaned by matching the raw values with the 48 official destinct values according to the dataset documentation using a text search with a distance matrix. An additional weather event category is introduced called “others” to group together all weather events that could not be matched with the official distinct values. This group accounts for around 10% of impact on crop damage (less for property damage, fatalities and injuries). Using the resulting clean data it was found that TORNADO is the most harmful weather event with respect to population health in the U.S. while FLOOD has the greates economic consequences.