01. Python y Pandas

Pandas summary

• The data table is a Pandas dataframe
• The data column is a Pandas series
• We can obtain information of the dataset:
  • 10 top rows: df.head(10)
  • Number of rows & variables: df.shape
  • Names of variables: df.columns
  • Types of variables: df.info()
  • Stats of numerical variables: df.describe()
  • Stats of numerical variables: df.describe(include=['object', 'bool'])
• Information of numerical variables:
  • Minimum df.num_var.min()
  • Maximun df.num_var.max()
  • Mean df.num_var.mean()
  • Median df.num_var.median()
  • Std df.variable.std()
• Information of categorial variables:
  • Unique values (count): df.cat_var.value_counts()
  • Unique values (percen): df.cat_var.value_counts(normalize=True)
  • 2 categorial vars (count): pd.crosstab(df.cat_var1, df.cat_var2, margins=True)
  • 2 categorial vars (percen): pd.crosstab(df.cat_var1, df.cat_var2, margins=True, normalize=True)
• Information of numerical & categorial variables:
  • Pivot table: df.pivot_table([num_var1, num_var2, ...], [cat_var1, cat_var2], aggfunc='mean')
• Fitering rows:
  • One filter: df[ condition ]
  • Multiple filters: df[ (condition1) & (condition2)]