09:58:44 In this data, the four field names under market have consistent suffixes: total_market_cap_usd 2324631531958.692 and total_volume_24h_usd 95482946596.28995 both end in usd; btc_dominance_pct 58.94 and market_cap_change_24h_pct -5.79 end in pct. But when I get to tickers, that consistency falls apart: BTCUSDT's quote_volume_24h 1985673314.094399, ETHUSDT 709598596.800239, BNBUSDT 129939201.67523, and SOLUSDT 302829967.03837 all use the same field name, with nothing at the end.
I considered automatically adding a column based on the suffix, but stopped when I got to these two entries: last 68283.47 and high_24h 70992.57. 24h only tells you the window, not the denomination. These are just personal records of my readings, not investment advice; suffixes only tell you so much about the units, and applying the rule blindly could mislabel fields that have no suffix. So I left the original field strings as they are and added a separate column for units, filled in by hand.
When you create tables, do you add a unit note for each column, or leave the whole table unparsed and rely only on the field names? I went with the latter, but still haven't settled on it.
I considered automatically adding a column based on the suffix, but stopped when I got to these two entries: last 68283.47 and high_24h 70992.57. 24h only tells you the window, not the denomination. These are just personal records of my readings, not investment advice; suffixes only tell you so much about the units, and applying the rule blindly could mislabel fields that have no suffix. So I left the original field strings as they are and added a separate column for units, filled in by hand.
When you create tables, do you add a unit note for each column, or leave the whole table unparsed and rely only on the field names? I went with the latter, but still haven't settled on it.