In [27]:

import pandas as pd
import plotly as pl
import altair as al

al.data_transformers.disable_max_rows() # remove 5k rows limit
al.themes.enable('dark') # set dark theme

Out[27]:

ThemeRegistry.enable('dark')

In [28]:

import yfinance as yf

def get_ticker_history(ticker, print_info = False):
    '''
    Retrieve Ticker history from yahoo finance.
    Argument ticker can be a list of tickers.
    '''
    if type(ticker)==list: # if list download data for every ticker.
        tickers = []
        for t in ticker:
            temp = get_ticker_history(t).reset_index()
            temp['ticker'] = t
            tickers.append(temp)
        return pd.concat(tickers, axis=0)

    ticke = yf.Ticker(ticker)

    # get stock info
    if print_info:
        print(ticke.info['shortName'])   

    # get historical market data
    hist = ticke.history(period="max")
    return hist

In [29]:

df = get_ticker_history(['SPY'])
df.to_csv('data.csv')
df.head(2)

Out[29]:

	Date	Open	High	Low	Close	Volume	Dividends	Stock Splits	ticker
0	1993-01-29 00:00:00-05:00	25.352047	25.352047	25.225917	25.334028	1003200	0.0	0	SPY
1	1993-02-01 00:00:00-05:00	25.352041	25.514208	25.352041	25.514208	480500	0.0	0	SPY

In [30]:

## Line Chart data
df['date'] = df.Date
df['month_day'] = [d.month*100 + d.day for d in df.date]
df['year'] = [str(d.year) for d in df.date]
df['year2'] = df.year

# calculate expanding pct change by year
df['ytd_pct_chng'] = df.groupby('year')['Close'].apply(lambda x: x.div(x.iloc[0]).subtract(1))
print(f'Original df {df.shape}:\n',df.head(2),'\n')

## Correlation Chart data
corrMat = df.pivot('month_day','year')['Close'].corr().stack()
corrMat.index.names = ["year", "year2"]
corrMat = corrMat.reset_index().rename(columns={0:'corr'})
corrMat = corrMat[~corrMat[['year','year2']].apply(frozenset, axis=1).duplicated()] # remove duplicate pairs
corrMat
print(f'Correlation df {corrMat.shape}:\n',corrMat.head(),'\n')

## Horizontal Bar data
# calculate percent change
annual = df.groupby('year')['Close'].apply(lambda x: pd.Series.pct_change(x).sum()).reset_index()
print(f'Annual Change df {annual.shape}:\n',annual.head(3))

Original df (7484, 14):
                        Date       Open       High        Low      Close  \
0 1993-01-29 00:00:00-05:00  25.352047  25.352047  25.225917  25.334028   
1 1993-02-01 00:00:00-05:00  25.352041  25.514208  25.352041  25.514208   

    Volume  Dividends  Stock Splits ticker                      date  \
0  1003200        0.0             0    SPY 1993-01-29 00:00:00-05:00   
1   480500        0.0             0    SPY 1993-02-01 00:00:00-05:00   

   month_day  year year2  ytd_pct_chng  
0        129  1993  1993      0.000000  
1        201  1993  1993      0.007112   

Correlation df (465, 3):
    year year2      corr
0  1993  1993  1.000000
1  1993  1994  0.497354
2  1993  1995  0.878284
3  1993  1996  0.826792
4  1993  1997  0.794767 

Annual Change df (30, 2):
    year     Close
0  1993  0.087365
1  1994  0.012181
2  1995  0.321471

In [31]:

## Line Chart
# define base chart
base = al.Chart(df,title="Every Year of SPY, 1993 - 2022").mark_line(interpolate='basis').encode(
    x=al.X('monthdate(date):O', title='',  axis=al.Axis(labelAngle=-45)),
    y=al.Y('ytd_pct_chng:Q', title='Percent Change YTD', axis=al.Axis(format='%')),
    detail='year',
    color = al.condition("datum.year == '2022'", al.value('red'), al.value('grey')),
    tooltip=['date',al.Tooltip('ytd_pct_chng', format=".0%")]
).properties(
    width=1100,
    height=500
)

# add highlight on hover selector
highlight = al.selection(type='single', on='mouseover',
                          fields=['year'], nearest=True)

points = base.mark_circle().encode(
    opacity=al.value(0)
).add_selection(
    highlight
)

# add hover highlight
lines = base.mark_line().encode(
    size=al.condition(~highlight, al.value(1), al.value(3))
)
# add 2022 color fill
line22 = base.mark_line().encode(
    size=al.condition(al.expr.datum['year'] == '2022', al.value(2), al.value(1))
)

spy_line = (points + lines + line22)
spy_line

Out[31]: