Learn_EDA_for_Data_Science

Univariate, Bivariate and Multi-variate Analysis

Data structure

Data Type Conversion

• coerce will introduce NA values for non numeric data in the columns
• if there are values that cannot be changed into numeric it will throw an error therefore the above statement

Remove Duplicates

• Count of Duplicated Rows
• print the duplicated rows
• Drop Columns
• Rename the weird columns

Outlier Detection

• Box plot
• Extracting Outliers
• Fliers are Outliers
• To get Whiskers

Descriptive Stats

Check for Balaced or Imbalanced Data in Categorical data

• Bar Plot

Missing Values and Imputation

• Mean Imputation

Null values Imputation for categorical data/values

• Get the object values
• Missing value imputation for categorical value
• Join the data set with imputed object dataset

Scatter plot and Correlation Analysis

Transformation of Data

• Creating Dummy Values for weather column

Normalization of the Data range(0 to 1)

• Summarize Transform data

Standardize data (0 mean, 1 std) range(-3 sigma to +3 sigma)

Speed Up EDA Process