Wasteman
@wasteman-codes.bsky.social
Overall, I don’t think BigDecimal is a bad option. Depending on your use case and organization, it may work perfectly fine.
My experience has just lead me to prefer BigInts. This is my recommended “Best Practice” but it’s not the only valid practice.
My experience has just lead me to prefer BigInts. This is my recommended “Best Practice” but it’s not the only valid practice.
July 15, 2024 at 5:47 AM
Overall, I don’t think BigDecimal is a bad option. Depending on your use case and organization, it may work perfectly fine.
My experience has just lead me to prefer BigInts. This is my recommended “Best Practice” but it’s not the only valid practice.
My experience has just lead me to prefer BigInts. This is my recommended “Best Practice” but it’s not the only valid practice.
E.g we had a situation when an ML team pulled some financial data into spark, and processed it as a FloatType instead of a BigDecimal. For obvious reasons, aggregations over the float value resulted in incoherent results.
July 15, 2024 at 5:45 AM
E.g we had a situation when an ML team pulled some financial data into spark, and processed it as a FloatType instead of a BigDecimal. For obvious reasons, aggregations over the float value resulted in incoherent results.
Some people make the argument that any team pulling financial data should know how to store it correctly. I agree, but my experience working at large tech companies is that this just isn’t the case.
July 15, 2024 at 5:44 AM
Some people make the argument that any team pulling financial data should know how to store it correctly. I agree, but my experience working at large tech companies is that this just isn’t the case.
Different languages & data-stores all have different ways of storing BigDecimal, and some don't even have a BigDecimal equivalent. But every language & data-store I know of stores a BigInt type.
It's easy for a team to pull your data into their system as an incompatible type. Less likely with Ints
It's easy for a team to pull your data into their system as an incompatible type. Less likely with Ints
July 15, 2024 at 5:42 AM
Different languages & data-stores all have different ways of storing BigDecimal, and some don't even have a BigDecimal equivalent. But every language & data-store I know of stores a BigInt type.
It's easy for a team to pull your data into their system as an incompatible type. Less likely with Ints
It's easy for a team to pull your data into their system as an incompatible type. Less likely with Ints
I only have 5/100 which I wasn't surprised about, the vast majority of the books I read don't appear on NYTimes best sellers
July 12, 2024 at 7:04 PM
I only have 5/100 which I wasn't surprised about, the vast majority of the books I read don't appear on NYTimes best sellers