How many measurements of historical data should I use for a reliable MACD result

I'm pretty new in this area. I have some historical data with the following columns ('Timestamp', 'Open', 'High', 'Low', 'Close', 'Volume').

Here is an example of the data

​

open high low close volume macd macds macdh

Timestamp

1558656000000 0.000008 0.000008 0.000008 0.000008 6097548.0 0.000000e+00 0.000000e+00 0.000000e+00

1558742400000 0.000008 0.000008 0.000008 0.000008 7469081.0 2.467949e-09 1.371083e-09 1.096866e-09

1558828800000 0.000008 0.000008 0.000008 0.000008 5375146.0 -3.677530e-09 -6.980210e-10 -2.979509e-09

1558915200000 0.000008 0.000008 0.000008 0.000008 5448930.0 -1.688887e-09 -1.033680e-09 -6.552069e-10

1559001600000 0.000008 0.000008 0.000008 0.000008 5725232.0 -4.154040e-09 -1.961917e-09 -2.192123e-09

… … … … … … … … …

1601424000000 0.000002 0.000002 0.000002 0.000002 8777722.0 -2.797820e-07 -2.532551e-07 -2.652694e-08

1601510400000 0.000002 0.000002 0.000002 0.000002 13569157.0 -2.830878e-07 -2.592216e-07 -2.386622e-08

1601596800000 0.000002 0.000002 0.000002 0.000002 13959722.0 -2.856427e-07 -2.645058e-07 -2.113686e-08

1601683200000 0.000002 0.000002 0.000002 0.000002 2694613.0 -2.843892e-07 -2.684825e-07 -1.590667e-08

1601769600000 0.000002 0.000002 0.000002 0.000002 1676424.0 -2.817616e-07 -2.711383e-07 -1.062328e-08

[500 rows x 8 columns]

​

As you can see I have 500 rows of data now. I've fetched this ohlcv data with a timestamp of 1 day. So this is data of 500 days. (Epoch timestamps differences 1 day).

​

Now as you can see I have the macd and macds data. Now I'm wondering how many values of a stock I need to get a valid macd result. I would assume that it's not a clean result if the measurements only start 2 days ago. Is there some algorithm to calculate how many rows of historical data you need to measure to result in a valid/reliable macd value? Does this differ for historical data of 1 day and for historical data of 1 minute (e.g. 1000 rows of 1 minute values)?