Graphing Data Set

**Pandas allows the user to use its own methods to graph plots in correspondance with Matplotlib. This allows users to pass arguements native to matplotlib through functions called through pandas**

Matplotlib

A python library used for the visualization of data

Used pyplot() to shows each graph to prevent overlapping data visualizations
Used ticker() to reformat the y axis ticker to visualize the greater numbers in numerical format and not decimal format with e

Seaborn

A python library used for data visualization which is based off of matplotlib

Seaborn was the preferred graphing library as it proves to be more beneficial in the visual aspect, as well as graphical skills
Seaborn regplot() was used to graph the correlation between total cases and total deaths
- The regression line was outlined with a 95% confidence interval
  - Unable to return the line of regression with its coeficient or intercept, therefore, scikit-learn was used to supplement correlational analysis

Folium

A python library used for data visualization on maps which is based off of leaflet.js

Folium was used for its abilites to plot accurate world maps
Folium Choropleth() was used to visualize total cases in each state in the United States of America
- Utilized a gradient color schemata to depict total number of positive Covid-19 cases
- Implemented Folium Marker() to display data upon user click, corresponding to the cursors lattitude and longitude position on the map
  - Identifying the state and its respective Covid-19 data: total cases and deaths and increase in cases and deaths on a specfied date

Scikit-learn

A python library focues on machine learning allowing the user to use various algorithms for predictive data analysis

Used LinearRegression() to calculate the intercept and coefficient of regression
- To show and analyze the correlational coefficent alongside the proportion of the variance
Sklearn.preprocessing => PolynomialFeatures(): establishing a matrix of the polynomial combinations of an entity based from the degree
- Create a polynomial like regression ==> establish the degree of the polynomial structure wanted in correlation analysis
Sklearn.pipeline => Pipeline(): used to transform a list based from given estimators
- Alter Linear regression to have polynomial like structure effectively giving correlation with respect inflection points in data set
  - Taking into account the second derivative of dataset presenting with fluctuations ==> overall postive slope that increases and decreases in magnitude
Using machine learning library allows constant computation of future data based from current data

Page updated

Google Sites

Report abuse