From 71f9f4a6743e740affa7ecc83af69f7b5191ddfc Mon Sep 17 00:00:00 2001
From: Palash Tyagi <23239946+Magnus167@users.noreply.github.com>
Date: Sat, 15 Mar 2025 00:03:09 +0000
Subject: [PATCH] Add logging for historic volatility calculations and
 implement utility to get min/max real dates

---
 src/panel/historic_vol.rs     | 19 ++++++++++++++++---
 src/panel/linear_composite.rs |  0
 src/utils/misc.rs             | 30 ++++++++++++++++++++++++++++++
 3 files changed, 46 insertions(+), 3 deletions(-)
 create mode 100644 src/panel/linear_composite.rs

diff --git a/src/panel/historic_vol.rs b/src/panel/historic_vol.rs
index 8b2baa2..01dcbe2 100644
--- a/src/panel/historic_vol.rs
+++ b/src/panel/historic_vol.rs
@@ -131,9 +131,15 @@ fn freq_period_calc(
         );
     }
 
+    println!("Calculating historic volatility with the following parameters:");
+    println!("lback_periods: {:?}, lback_method: {:?}, half_life: {:?}, remove_zeros: {:?}, nan_tolerance: {:?}, period: {:?}", lback_periods, lback_method, half_life, remove_zeros, nan_tolerance, period);
+
     let mut new_df = dfw.clone();
 
     for col_name in dfw.get_column_names() {
+        if col_name == "real_date" {
+            continue;
+        }
         let series = dfw.column(col_name)?;
         let values: Array1<f64> = series
             .f64()?
@@ -174,7 +180,7 @@ fn freq_period_calc(
             }
             _ => return Err("Invalid lookback method.".into()),
         };
-
+        println!("Successfully calculated result_series for column: {:?}", col_name);
         new_df.with_column(result_series)?;
     }
 
@@ -298,9 +304,14 @@ pub fn historic_vol(
 
     println!("Successfully got nan_tolerance.");
 
-    let start_date = NaiveDate::parse_from_str(&start, "%Y-%m-%d")?;
-    let end_date = NaiveDate::parse_from_str(&end, "%Y-%m-%d")?;
+    let (dfw_start_date, dfw_end_date) =
+        crate::utils::misc::get_min_max_real_dates(&dfw, "real_date")?;
+    println!("Successfully got min and max real dates.");
 
+    let (start_date, end_date) = (
+        NaiveDate::parse_from_str(&start, "%Y-%m-%d").unwrap_or_else(|_| dfw_start_date),
+        NaiveDate::parse_from_str(&end, "%Y-%m-%d").unwrap_or_else(|_| dfw_end_date),
+    );
     println!("Successfully parsed start and end dates.");
 
     dfw = dfw
@@ -327,6 +338,8 @@ pub fn historic_vol(
         _ => return Err("Invalid frequency specified.".into()),
     };
 
+    println!("Successfully got period.");
+
     let dfw = match est_freq.as_str() {
         "D" => freq_daily_calc(
             &dfw,
diff --git a/src/panel/linear_composite.rs b/src/panel/linear_composite.rs
new file mode 100644
index 0000000..e69de29
diff --git a/src/utils/misc.rs b/src/utils/misc.rs
index c664708..678062f 100644
--- a/src/utils/misc.rs
+++ b/src/utils/misc.rs
@@ -1,3 +1,4 @@
+use chrono::NaiveDate;
 use polars::prelude::*;
 use std::collections::HashMap;
 use std::error::Error;
@@ -11,6 +12,35 @@ pub fn split_ticker(ticker: String) -> Result<(String, String), Box<dyn Error>>
     Ok((parts[0].to_string(), parts[1].to_string()))
 }
 
+pub fn get_min_max_real_dates(
+    df: &DataFrame,
+    date_col: &str,
+) -> Result<(NaiveDate, NaiveDate), Box<dyn Error>> {
+    let date_series = df.column(date_col)?;
+    if let DataType::Date = date_series.dtype() {
+        // Convert the `date` series to an i32 (days since 1970-01-01)
+        let date_as_days = date_series.cast(&DataType::Int32)?;
+        let min_days = date_as_days.i32()?.min().ok_or("No minimum value found")?;
+        let max_days = date_as_days.i32()?.max().ok_or("No maximum value found")?;
+
+        // Convert the days back to `NaiveDate`
+        let min_date = NaiveDate::from_ymd_opt(1970, 1, 1)
+            .unwrap()
+            .checked_add_signed(chrono::Duration::days(min_days as i64))
+            .ok_or("Invalid minimum date")?;
+        let max_date = NaiveDate::from_ymd_opt(1970, 1, 1)
+            .unwrap()
+            .checked_add_signed(chrono::Duration::days(max_days as i64))
+            .ok_or("Invalid maximum date")?;
+
+        Ok((min_date, max_date))
+    } else {
+        Err(Box::new(polars::error::PolarsError::ComputeError(
+            "The column is not of Date type".into(),
+        )))
+    }
+}
+
 #[allow(dead_code)]
 pub fn get_cid(ticker: String) -> Result<String, Box<dyn Error>> {
     split_ticker(ticker).map(|(cid, _)| cid)