TradingView VWAP-Stdev-Bands-v2-Mod-UPDATE in Python

data.csv

Date	Open	High	Low	Close	Volume
2020-05-06 18:15:00-04:00	2832	2838.5	2832	2838	3750
2020-05-06 18:30:00-04:00	2838	2838.5	2826	2827.75	6864
2020-05-06 18:45:00-04:00	2827.75	2830.5	2827	2829.5	2937
2020-05-06 19:00:00-04:00	2829.5	2830.5	2823	2826.5	4619
2020-05-06 19:15:00-04:00	2826.75	2829.5	2825.25	2827.75	3610
2020-05-06 19:30:00-04:00	2827.75	2829.5	2825.5	2826.75	2460
2020-05-06 19:45:00-04:00	2826.75	2830.75	2826.25	2830.5	2531
2020-05-06 20:00:00-04:00	2830.5	2833.5	2830.25	2833.5	2361

the codez

import pandas as pd
import mplfinance as mpf
import math

df = pd.read_csv('./data.csv', sep=',', quotechar='"')

df.set_index(['Date'], inplace=True)
df.index = pd.to_datetime(df.index)
df.index.name = 'Date'

# from here = https://www.tradingview.com/script/fD2ppI7I-VWAP-Stdev-Bands-v2-Mod-UPDATE/
# from tradingview = hl2 = midpoint of bar
df['HL2'] = (df.High + df.Low) / 2
df['VOLUME_SUM'] = df.Volume.cumsum()
df['VWAP'] = (df.Volume * df.HL2).cumsum() / df.VOLUME_SUM
# y'all doing some weird ass variance calculation here
df['V2SUM'] = (df.Volume * df.HL2 * df.HL2).cumsum()

def sigma_tradingview(x):
    val = x.V2SUM / x.VOLUME_SUM - x.VWAP * x.VWAP
    # kind of a hack to mimic that tv max() function
    if val <= 0:
        val = 0
    return math.sqrt(val)

df['STDEV_TV'] = df.apply(sigma_tradingview, axis=1)

stdev_multiple_1 = 1.28
stdev_multiple_2 = 2.01
stdev_multiple_3 = 2.51

df['STDEV_1'] = df.VWAP + stdev_multiple_1 * df['STDEV_TV']
df['STDEV_N1'] = df.VWAP - stdev_multiple_1 * df['STDEV_TV']

addplot  = [
    mpf.make_addplot(df['VWAP']),
    mpf.make_addplot(df['STDEV_1']),
    mpf.make_addplot(df['STDEV_N1']),
]

mpf.plot(df, type='candle', addplot=addplot)

Blue = VWAP
Orange = VWAP + 1 dev
Green = VWAP - 1 dev

TradingView

How do I write vwap z score code