pandas.md

Understanding Datasets

Function	Description
dataframe.head(n)	to show the first n rows of dataframe || else top 5
dataframe.tail(n)	to show the bottom n rows of dataframe || else bottom 5
dataframe.shape	size of dataframe (rows, columns)
dataframe.dtypes()	to check data types
dataframe.describe()	returns a statistical summary
dataframe.describe(include="all")	returns a statistical summary of all including objects
dataframe.info()	provides a concise summary of all your dataframe
dataframe.astype()	to convert data types
dataframe.columns.values	to get a array of column headers from
dataframe.index.values	to get a array of index

Dataframe Manipulation

Command	Description
dataframe.set_index("key", inplace=True)	changes the index of the dataframe to given column "key"
dataframe.reset_index()	resets the index of the dataframe
dataframe.dropna(axis=0)	drops the entire row if found empty
dataframe.dropna(axis=1)	drops the entier column if found empty
dataframe.dropna( [keys] ,inplace=True)	applies changes to the actual dataframe itself
dataframe.replace( missing_value, new_value )	replace value of one to another in dataframe
dataframe.rename( columns={prevkey:newkey} )	rename columns of the dataframe
dataframe.sum( axis=1 )	returnsm sum of the row
dataframe["key"].std()	returns the standard deviation
dataframe["key"].mean()	returns the mean of the column
pd.get_dummies(df[column])	Convert categorical variables to dummy variables (0 or 1) [One - hot encoding]
dataframe.groupby([column, column])	Group dataframe based on column
dataframe[column].agg('function')	Perform an operation in a particular coulmn

Viewing Datasets

Command	Description
dataframe.loc[label]	filters by the labels of the index/column # For val of particular col and row use dataframe.loc['label','col']
dataframe.iloc[index]	filters by the position of the index/coumn

Miscellaneous

Data Format	Read	Save
csv	pd.read_csv()	df.to_csv()
json	pd.read_json()	df.to_json()
Excel	pd.read_excel()	df.to_excel()
sql	pd.read_sql()	df.to_sql()