From: Erik Mackdanz <erikmack@gmail.com>
Date: Sun, 26 Nov 2023 14:54:15 +0000 (-0600)
Subject: convert eager df to lazy df
X-Git-Url: https://git.humopery.space/?a=commitdiff_plain;h=672ff1ec7e92d000ee25ed2ad169a74356d01ae9;p=polarsbop.git

convert eager df to lazy df
---

diff --git a/src/main.rs b/src/main.rs
index 89c652d..cec9664 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -24,7 +24,7 @@ async fn main() -> Result<(),PolarsError> {
     let mut schema_override = Schema::new();
     let _ = schema_override.with_column("2022".into(),DataType::Float64);
     let schema_override =
-	Some(Arc::new(schema_override));
+	Some(&schema_override);
 
     // Some string values in numeric fields should be ignored
     let null_values =
@@ -40,38 +40,40 @@ async fn main() -> Result<(),PolarsError> {
 	)));
 
     // read and parse from disk
-    let mut df = CsvReader::from_path("data/bop.csv")?
+    let mut df = LazyCsvReader::new("data/bop.csv")
 	.has_header(true)
-	.with_dtypes(schema_override)
+	.with_dtype_overwrite(schema_override)
 	.with_null_values(null_values)
 	.finish()?;
 
     // split list based on arg
     match partition.as_str() {
 	"0" => {
-	    let mask = df.column("Country Name")?.lt("Ethiopia")?;
-	    df = df.filter(&mask)?;
+	    let mask = col("Country Name").lt(lit("Ethiopia"));
+	    df = df.filter(mask);
 	},
 	"1" => {
-	    let mask = df.column("Country Name")?.gt_eq("Ethiopia")?;
-	    df = df.filter(&mask)?;
-	    let mask = df.column("Country Name")?.lt("Romania")?;
-	    df = df.filter(&mask)?;
+	    let mask = col("Country Name").gt_eq(lit("Ethiopia"));
+	    df = df.filter(mask);
+	    let mask = col("Country Name").lt(lit("Romania"));
+	    df = df.filter(mask);
 	},
 	"2" => {
-	    let mask = df.column("Country Name")?.gt_eq("Romania")?;
-	    df = df.filter(&mask)?;
+	    let mask = col("Country Name").gt_eq(lit("Romania"));
+	    df = df.filter(mask);
 	},
 	_ => { panic!("partition must be 0, 1 or 2"); }
     }
 
     // values not status, selected cols
     let mut df = df
-	.filter(&df.column("Attribute")?.
-			   equal("Value")?)?
-	.select(["Country Name","Indicator Name","Indicator Code","2022"])?
-	.rename("2022","Value")?
-	.sort(["Country Name","Indicator Name"],vec!(false,false),false)?;
+	.filter(col("Attribute").eq(lit("Value")))
+	.select(&[col("Country Name"),col("Indicator Name"),
+		  col("Indicator Code"),col("2022")])
+	.rename(["2022"],["Value"])
+	.sort_by_exprs([col("Country Name"),col("Indicator Name")],
+		       vec!(false,false),false,false)
+	.collect()?;
 
     // some console output
     println!("df: {:?}",df);