databyjp

# See sample charts of correlated stocks
sim_charts = np.argsort(r_array[0])[-6:]
sim_syms = [filt_df.sym.unique()[i] for i in sim_charts]
sim_df = filt_df[filt_df.sym.isin(sim_syms)]
fig = px.scatter(sim_df,
                 x='date', y='close', color='sym', template='plotly_white', title='Sample Stock Price Graphs',
                 facet_col='sym', facet_col_wrap=3
                 )
fig.update_yaxes(matches=None, showticklabels=False)  # If we want to simply view curve 'shapes'
fig.update_traces(mode='lines')

databyjp

# Cluster data
model = sklearn.cluster.SpectralBiclustering(n_clusters=3, method='log', random_state=0)
model.fit(r_array)

r_array_srt = r_array[np.argsort(model.row_labels_)]
r_array_srt = r_array_srt[:, np.argsort(model.column_labels_)]
syms_srt = filt_df.sym.unique()[np.argsort(model.row_labels_)]

fig = px.imshow(r_array_srt, x=syms_srt, y=syms_srt,
                color_continuous_scale=px.colors.sequential.Bluyl, color_continuous_midpoint=0,

databyjp

fig = px.imshow(r_array, x=symbols_ser, y=symbols_ser,
                color_continuous_scale=px.colors.sequential.Bluyl, color_continuous_midpoint=0,
                )
fig.update_xaxes(side="top")
fig.show()

databyjp

>>> print(r_array)
[[ 1.          0.94783247  0.75459535  0.89186946  0.93008958 -0.68727618]
 [ 0.94783247  1.          0.79397608  0.91511258  0.92275757 -0.71430629]
 [ 0.75459535  0.79397608  1.          0.74710565  0.87785573 -0.3759863 ]
 [ 0.89186946  0.91511258  0.74710565  1.          0.86919002 -0.61902468]
 [ 0.93008958  0.92275757  0.87785573  0.86919002  1.         -0.5900107 ]
 [-0.68727618 -0.71430629 -0.3759863  -0.61902468 -0.5900107   1.        ]]

databyjp

# Build a correlation (and p-value) matrix
r_array = np.zeros([len(symbols_ser), len(symbols_ser)])
p_array = np.zeros([len(symbols_ser), len(symbols_ser)])
for i in range(len(symbols_ser)):
    for j in range(len(symbols_ser)):
        ser_i = total_ticker_df[total_ticker_df.sym == symbols_ser[i]]['close'].values
        ser_j = total_ticker_df[total_ticker_df.sym == symbols_ser[j]]['close'].values
        r_ij, p_ij = scipy.stats.pearsonr(ser_i, ser_j)
        r_array[i, j] = r_ij
        p_array[i, j] = p_ij

databyjp

fig = px.scatter(total_ticker_df,
                 x='date', y='close', color='sym', template='plotly_white', title='Sample Stock Price Graphs',
                 facet_col='sym', facet_col_wrap=3
                 )
fig.update_traces(mode='lines')
fig.show()

databyjp

# ========== FUNCTION TO GET STOCK DATA ==========
def get_stock_data(tkn, sym='amzn', start_date='2020-01-01'):
    headers = {'Content-Type': 'application/json'}

    requestResponse = requests.get("https://api.tiingo.com/tiingo/daily/" + sym + "/prices?startDate=" + start_date + "&token=" + tkn, headers=headers)
    if requestResponse.status_code == 200:
        logger.info(f'Success fetching {sym} data from {start_date} to today')
    else:
        logger.warning(f'Something looks wrong - status code {requestResponse.status_code}')

databyjp

    card_content_worth = [
        dbc.CardHeader(f"Bezos' net worth has increased by: "),
        dbc.CardBody([
            html.H3([dbc.Badge(f"US${int(bezos_secondly * n)}", color='success')], className="card-title"),
            html.P([
                f"That is ", dbc.Badge(str(round(bezos_tot/median_us_income, 1)), color='primary'),
                " typical U.S. incomes' worth of money."],
                className="card-text",
            ),
        ])]

databyjp

def get_bezos_data():
    amzn_data = get_stock_data(tiingo_token).json()
    amzn_df = pd.DataFrame(amzn_data)
    amzn_df = amzn_df.assign(bezos_year=(amzn_df.close - amzn_df.iloc[0].close) * 498 * 10**6 * 0.112)
    return amzn_df

app.layout = dbc.Container([
    html.Div([
        dcc.Graph(
            id='bezos-worth-graph',

databyjp

app.layout = html.Div([
    html.Div([
        dcc.Graph(id='graph_id'),
        dcc.Interval(
            id='interval_id',
            interval=1 * 1000,  # in milliseconds
            n_intervals=0
        )
    ]),
])