CODEWITHSUNDEEP
pythondataframepython-datetime
Patrizio
Patrizio

Reputation: 3

Iteration column with Day's name

I Would like to Compare the Nasdaq's Monday's daily performance with the PREVIOUS Friday's performance, if they are the same record event in a counter. I create a column that specifies if Nasdaq went Down (RIBASSO) or up (RIALZO), then i think to extract only Monday and Friday rows. The desired output is a counter that record every time this event occur.

import yfinance as yf
import pandas as pd
import numpy as np

database = yf.download('NQ=F')
df = pd.DataFrame(database)

# Day's name column
df["Day's name"] = df.index.to_series().dt.day_name()
df.drop(columns= ['Adj Close', 'Volume'])

# Daily performance column 
conditions = [ df['Close'] - df['Open'] > 0,
             df['Close'] - df['Open'] < 0,
             df['Close'] - df['Open'] == 0]
categories = ['RIALZO', 'RIBASS0', 'FLAT']
df['Daily Performance'] = np.select(conditions, categories, default='Unknown')


# Monday/Friday Musk
monday_perf = df["Day's name"] == 'Monday' 
friday_perf = df["Day's name"] == 'Friday'
df_mask = df[monday_perf | friday_perf]

Upvotes: 0

Views: 32

Answers (1)

sud
sud

Reputation: 165

Use shift along with cumsum to achieve the desired results.

import yfinance as yf
import pandas as pd
import numpy as np

database = yf.download('NQ=F')
df = pd.DataFrame(database)

# Day's name column
df["Day's name"] = df.index.to_series().dt.day_name()
df.drop(columns= ['Adj Close', 'Volume'])

# Daily performance column 
conditions = [ df['Close'] - df['Open'] > 0,
             df['Close'] - df['Open'] < 0,
             df['Close'] - df['Open'] == 0]
categories = ['RIALZO', 'RIBASS0', 'FLAT']
df['Daily Performance'] = np.select(conditions, categories, default='Unknown')

# shift daily performance by 1 to get previous daily performance
df['Previous Daily Performance'] = df['Daily Performance'].shift(1)
# assign 1/0 to cases where current matches previous
df['Monday Equals Previous Friday'] = df[['Daily Performance', 'Previous Daily Performance']] \
    .apply(lambda x: 1 if x[0] == x[1] else 0, axis=1)

# Keep only Monday rows since previous Friday value is available in a separate column
df_mondays = df[df["Day's name"] == 'Monday']

# add a counter to count instances of match
df_mondays['Counter'] = df_mondays['Monday Equals Previous Friday'].cumsum()

# use this to see the total count of matches
print(df_mondays['Counter'].max())

Upvotes: 0

Related Questions