Pandas

Pandas

Data Frame

• Importing Pandas Module import pandas as pd

• Reading CSV File in Pandas df = pd.read_csv(, sep=",")

• Showing top Records - head()

• Showing last Records - tail()

• Shape of Data Frame

• Data Frame size (rows*columns)

• Number of Rows in Data Frame - len()

• Column names in Data Frame

	# Print All Columns 
	print(df.columns) 
	
	# iterating the columns 
	for col in df.columns: 
    	print(col) 

• Exctracting Data from Pandas

  1. Get Single Column

  2. Get Range of rows in Single Column

  3. Get Single value in Column

  4. Get Multiple Columns

• Sort Data Frame

• Count Values in Column

• Plot a Data Frame

• Replace values in Column

  df[<Column Name>].replace(<Value to Serch>,<Value to Replace>,inplace=True)

  df['Gender'] = df['Gender'].map({'F': 'Female', 'M': 'Male'})

• Printing Column in Pandas print(df[])

• Rename Columns in Pandas df.columns = ['Column Name', 'Column Name', 'Column Name', 'Column Name']

   df = df.rename(columns={"A": "a", "B": "c"})

• Unique Values In A Column print(df..unique())

• Append a column in Dataframe

	df[<Column Name>]=<Value to be appended>

• Merge Columns in Dataframe

	df['ColumnA'] = df[df.columns[1:]].apply(
   lambda x: ','.join(x.dropna().astype(str)),
   axis=1)

• Insert a column DataFrameName.insert(loc, column, value, allow_duplicates = False)

• Remove a column

	df = df.drop(columns="Column Name")

Pandas Plot

• Plotting a Dataframe

import pandas as pd import matplotlib.pyplot as plt df = pd.DataFrame({ 'name':['john','mary','peter','jeff','bill','lisa','jose'], 'age':[23,78,22,19,45,33,20] }) df.plot() plt.show()

- Showing legends
````python
 plt.legend(loc='best')

• Adding Labels to Plot lines

	df.plot(label="AGE")

• Convert column to Date time format

  df['Month'] = pd.to_datetime(df['Month'],infer_datetime_format=True)

• Keep Most Recent date value in Dataframe

• Convert API Response to Dataframe

response = requests.request("POST", URL, headers={'Content-Type': 'application/json'}, data=json_input)
json_output = json.loads(response.text)
df = json_normalize(json_output)
df.to_csv("data\ddf.csv", sep=',', index=False)

Merge Dataframes

Slice Dataframe based on Column Value