Make DataFrame inferred type in compile-time

[Kotlin-GDE]

when using dataframe without kotlin notbook, columns not inferred in compile-time
how to fix that ?

[Jolanrensen]

Hi there :)

For columns to be inferred in your normal Kotlin project, you will need the Compiler plugin. This will provide column accessors for any DataFrame<YourType> where YourType is a data schema interface/class you defined. (You might need to cast your dataframe to the data schema type.)
After running any operation, like add, remove, etc. the column accessors will be updated on the fly :).

If you are using an older version of DataFrame, Kotlin, or IntelliJ and you cannot yet use the compiler plugin, we also have an older Gradle/KSP plugin which can also generate column accessors, but those won't be updated on the fly after running operations. So it can only generate accessors when a data schema is defined.

Hope this can get you started!

[linux-china]

@Jolanrensen do you have any example on GitHub for Compiler plugin? I could not find Kotlin DataFrame compiler plugin example from dataframe repository.

[Kotlin-GDE]

hi, from docs

plugins {
    id("org.jetbrains.kotlinx.dataframe") version "1.0.0-Beta2"
}

dependencies {
    implementation("org.jetbrains.kotlinx:dataframe:1.0.0-Beta2")
}

plugin not compatible with kotlin 2.1.20 !!

[Jolanrensen]

@linux-china You're right, we didn't put an example there yet, we'll certainly add it soon, once everything is a bit more stable :) (we do have an older example project here but it's out-of-date)

In the meantime @Kotlin-GDE is right, following https://kotlin.github.io/dataframe/compiler-plugin.html#setup should get you started (if you exactly follow each step of course), this means having the right IDE version, Kotlin version, etc.