numerai-tools 0.5.0.dev9__tar.gz → 0.5.0.dev11__tar.gz

{numerai_tools-0.5.0.dev9 → numerai_tools-0.5.0.dev11}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: numerai-tools
-Version: 0.5.0.dev9
+Version: 0.5.0.dev11
 Summary: A collection of open-source tools to help interact with Numerai, model data, and automate submissions.
 License: MIT
 Author: Numerai Engineering

{numerai_tools-0.5.0.dev9 → numerai_tools-0.5.0.dev11}/numerai_tools/scoring.py

@@ -560,7 +560,7 @@ def generate_neutralized_weights(
     neutralizers: pd.DataFrame,
     sample_weights: pd.Series,
     center_and_normalize: bool = False,
-) -> pd.Series:
+) -> pd.DataFrame:
     assert not predictions.isna().any().any(), "Predictions contain NaNs"
     assert not neutralizers.isna().any().any(), "Normalization factors contain NaNs"
     assert not sample_weights.isna().any(), "Weights contain NaNs"

{numerai_tools-0.5.0.dev9 → numerai_tools-0.5.0.dev11}/numerai_tools/signals.py

@@ -76,106 +76,92 @@ def turnover(
 
 
 def calculate_max_churn_and_turnover(
-    curr_sub: pd.DataFrame,
+    curr_sub: pd.Series,
     curr_neutralizer: pd.DataFrame,
-    curr_weight: pd.Series,
-    prev_week_subs: dict[str, pd.DataFrame],
+    curr_sample_weight: pd.Series,
+    prev_subs: dict[str, pd.Series],
     prev_neutralizers: dict[str, pd.DataFrame],
     prev_sample_weights: dict[str, pd.Series],
-    universe: pd.DataFrame,
-    curr_signal_col: str,
-    curr_ticker_col: str,
 ) -> Tuple[float, float]:
     """Calculate the maximum churn and turnover with respect to previous submissions.
+    This function iterates over previous submissions and calculates churn and turnover
+    for each submission against the current submission. It expects all data to be
+    indexed on the same type tickers/IDs (e.g. all numerai_ticker, or all composite_figi, or all etc.) .
 
     Arguments:
-        curr_sub -- the current submission
-        curr_neutralizer -- the neutralizer DataFrame for the current submission
-        curr_weight -- the sample weights Series for the current submission
-        prev_week_subs -- a dictionary of datestamps to submissions
-        prev_neutralizers -- a dictionary of datestamps to neutralizers
-        prev_sample_weights -- a dictionary of datestamps to sample weights
-        universe -- the universe DataFrame for the current era
-        curr_signal_col -- the column name for signal in the current submission
-        curr_ticker_col -- the column name for tickers in the current submission
-
+        curr_sub: pd.Series - the current submission as a Series indexed on tickers/ids
+        curr_neutralizer: pd.DataFrame - the neutralizer DataFrame for the current submission indexed on numerai_ticker
+        curr_sample_weight: pd.Series - the sample weights Series for the current submission indexed on numerai_ticker
+        prev_subs: dict[str, pd.DataFrame] - a dictionary of datestamps to submissions, where each submission is a DataFrame
+                     with 2 columns: a ticker/id column and a signal/prediction column. To calculate churn
+                     and turnover for a live submission, use the most recent 5 submissions. For diagnostics,
+                     just provide the previous era.
+        prev_neutralizers: dict[str, pd.DataFrame] - a dictionary of datestamps to neutralizers DataFrames where each neutralizers
+                             DataFrame is indexed on the same ticker column as the current submission
+        prev_sample_weights: dict[str, pd.Series] - a dictionary of datestamps to sample weights where each sample weights
+                             Series is indexed on the same ticker column as the current submission
     Returns:
         prev_week_max_churn -- the maximum churn from previous submissions
         prev_week_max_turnover -- the maximum turnover from previous submissions
     """
-    universe = universe.reset_index()
     (
         curr_ticker_col,
         curr_signal_col,
         _,
-        curr_sub,
+        curr_sub_df,
         _,
     ) = validate_submission_signals(
-        universe=universe,
-        submission=curr_sub,
+        universe=curr_sample_weight.index.to_frame(),
+        submission=curr_sub.reset_index(),
     )
-    curr_sub_vector = clean_submission(
-        universe=universe,
-        submission=curr_sub,
+    curr_sub = clean_submission(
+        universe=curr_sample_weight.index.to_frame(),
+        submission=curr_sub_df,
         src_id_col=curr_ticker_col,
         src_signal_col=curr_signal_col,
         rank_and_fill=True,
     )
+    print("curr_sub", curr_sub)
     churn_stats = []
     turnover_stats = []
     neutralized_weights = generate_neutralized_weights(
-        curr_sub_vector.to_frame(),
+        curr_sub.to_frame(),
         curr_neutralizer,
-        curr_weight,
+        curr_sample_weight,
         center_and_normalize=True,
-    )
-    for datestamp in prev_week_subs:
-        prev_sub = prev_week_subs[datestamp]
+    )[curr_sub.name]
+    for datestamp in prev_subs:
+        prev_sub = prev_subs[datestamp]
         prev_neutralizer = prev_neutralizers[datestamp]
-        prev_weight = prev_sample_weights[datestamp]
+        prev_sample_weight = prev_sample_weights[datestamp]
         (
             prev_ticker_col,
             prev_signal_col,
             _,
-            prev_sub,
+            prev_sub_df,
             _,
         ) = validate_submission_signals(
-            universe=universe,
-            submission=prev_sub,
+            universe=prev_sample_weight.index.to_frame(),
+            submission=prev_sub.reset_index(),
         )
-        filtered_prev_sub = clean_submission(
-            universe=universe,
-            submission=prev_sub,
+        prev_sub = clean_submission(
+            universe=prev_sample_weight.index.to_frame(),
+            submission=prev_sub_df,
             src_id_col=prev_ticker_col,
             src_signal_col=prev_signal_col,
             dst_id_col=curr_ticker_col,
             dst_signal_col=curr_signal_col,
             rank_and_fill=True,
         )
-        prev_neutralizer = (
-            remap_ids(
-                prev_neutralizer.reset_index(),
-                universe,
-                str(prev_neutralizer.index.name),
-                curr_ticker_col,
-            )
-            .set_index(curr_ticker_col)
-            .filter(like="neutralizer_")
-        )
-        prev_weight = remap_ids(
-            prev_weight.reset_index(),
-            universe,
-            str(prev_weight.index.name),
-            curr_ticker_col,
-        ).set_index(curr_ticker_col)[prev_weight.name]
         prev_neutralized_weights = generate_neutralized_weights(
-            filtered_prev_sub.to_frame(),
+            prev_sub.to_frame(),
             prev_neutralizer,
-            prev_weight,
+            prev_sample_weight,
             center_and_normalize=True,
-        )
+        )[prev_sub.name]
+        print("prev_sub", prev_sub)
         try:
-            churn_val = abs(churn(curr_sub_vector, filtered_prev_sub))
+            churn_val = abs(churn(curr_sub, prev_sub))
         except AssertionError as e:
             if "does not have enough overlapping ids" in str(e):
                 continue

{numerai_tools-0.5.0.dev9 → numerai_tools-0.5.0.dev11}/numerai_tools/submissions.py

@@ -88,6 +88,7 @@ def validate_headers_signals(
         assert (
             date_col_name is not None
         ), "invalid_submission_headers: submission must contain a date column"
+    print(submission)
     ticker_col, signal_col = _validate_headers(
         submission,
         SIGNALS_ALLOWED_ID_COLS,
@@ -228,11 +229,8 @@ def validate_submission_signals(
             "data_type column found in Signals submission. This is deprecated and support will be removed in the future. "
             "Please remove the data_type column from your Signals submission."
         )
-        submission.drop(
-            columns=["data_type"],
-            errors="ignore",
-            inplace=True,
-        )
+    submission.drop(columns=["data_type"], errors="ignore", inplace=True)
+    print(submission)
     ticker_col, signal_col, date_col = validate_headers_signals(
         submission, assert_date_col
     )
@@ -266,7 +264,7 @@ def validate_submission_crypto(
 
 def remap_ids(
     data: pd.DataFrame,
-    ticker_map: pd.Series | pd.DataFrame,
+    ticker_map: pd.DataFrame,
     src_id_col: str,
     dst_id_col: str,
 ) -> pd.DataFrame:
@@ -277,25 +275,25 @@ def remap_ids(
 
     Arguments:
         data: pd.DataFrame - the data to remap
-        ticker_map: pd.Series | pd.DataFrame - the mapping of source ids to destination ids
+        ticker_map: pd.DataFrame - the mapping of source ids to destination ids
         src_id_col: str - the name of the source ids column in the data
         dst_id_col: str - the name of the destination ids column in the ticker map
     """
     # first, index the universe and data on the source ids
-    indexed_map = ticker_map.reset_index().set_index(src_id_col)
+    indexed_map = ticker_map.set_index(src_id_col, drop=False)
     indexed_data = data.set_index(src_id_col)
     return (
         # then, join the universe and data
         indexed_map.join(indexed_data)
         # get just the destination ids and prediction columns
-        .reset_index()[[dst_id_col, *indexed_data.columns]]
+        .reset_index(drop=True)[[dst_id_col, *indexed_data.columns]]
         # finally, sort by the destination ticker column
         .sort_values(dst_id_col)
     )
 
 
 def clean_submission(
-    universe: pd.Series | pd.DataFrame,
+    universe: pd.DataFrame,
     submission: pd.DataFrame,
     src_id_col: str,
     src_signal_col: str,

{numerai_tools-0.5.0.dev9 → numerai_tools-0.5.0.dev11}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "numerai-tools"
-version = "0.5.0.dev9"
+version = "0.5.0.dev11"
 description = "A collection of open-source tools to help interact with Numerai, model data, and automate submissions."
 authors = [
     {name = "Numerai Engineering",email = "engineering@numer.ai"}