RSI Calculation

kevin.girard

Hi,

I would like to know the RSI calculation used on all standard reuters charts. My issue is that I have tried the pandas_ta.rsi function as well as the talib RSI function for the ric 'LCOc1' and others. These values are completely differnt to the ones I see on the chart. please see below my code and output:

rd.open_session()

df = rd.get_history(
["LCOc1"],
interval="daily",
fields=["SETTLE"]
).tail(20)

df["RSI_pandas_ta"] = ta.rsi(df["SETTLE"], length=14, talib=True)
df["RSI_talib"] = tl.RSI(df["SETTLE"])

print(df)

code_RSI_output.png

for LCOc1 the RSI i see on the chart for March 20th is: 48.9574

I have tried the method on python from this article: https://developers.lseg.com/en/article-catalog/article/technical-analysis-indicators-customisable-relative-strength-indices

but this gives the same results as pandas_ta version.

It would be useful to see the source formulas you use, python implemented if possible.

Maybe the errors could be either number rounding from the API? or exchange timezone close differences?

Gurpreet

Hello @kevin.girard

While there are a lot of RSI fields provided in Workspace as shown in this question; we don't know the exact calculation used in the built in chart. I would suggest reaching out to helpdesk at LSEG MyAccount to ask this content related question.