Merge 3e6b73424b into 2131205db6

freqtrade/data/converter.py

@@ -122,6 +122,47 @@ def ohlcv_fill_up_missing_data(dataframe: DataFrame, timeframe: str, pair: str)
     return df
 
 
+def reduce_mem_usage(pair: str, df: DataFrame) -> DataFrame:
+    """ iterate through all the columns of a dataframe and modify the data type
+        to reduce memory usage.
+    """
+    # start_mem = df.memory_usage().sum() / 1024**2
+    # logger.info(f"Memory usage of dataframe for {pair} is {start_mem:.2f} MB")
+
+    for col in df.columns[1:]:
+        col_type = df[col].dtype
+
+        if col_type != object:
+            c_min = df[col].min()
+            c_max = df[col].max()
+            if str(col_type)[:3] == "int":
+                # if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:
+                #     df[col] = df[col].astype(np.int8)
+                # elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:
+                #     df[col] = df[col].astype(np.int16)
+                if c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:
+                    df[col] = df[col].astype(np.int32)
+                elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:
+                    df[col] = df[col].astype(np.int64)
+            elif str(col_type)[:5] == "float":
+                # if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:
+                #     df[col] = df[col].astype(np.float16)
+                if c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:
+                    df[col] = df[col].astype(np.float32)
+                else:
+                    df[col] = df[col].astype(np.float64)
+            # else:
+            #     logger.info(f"Column not optimized because the type is {str(col_type)}")
+        # else:
+            # df[col] = df[col].astype('category')
+
+    # end_mem = df.memory_usage().sum() / 1024**2
+    # logger.info("Memory usage after optimization is: {:.2f} MB".format(end_mem))
+    # logger.info("Decreased by {:.1f}%".format(100 * (start_mem - end_mem) / start_mem))
+
+    return df
+
+
 def trim_dataframe(df: DataFrame, timerange, df_date_col: str = 'date',
                    startup_candles: int = 0) -> DataFrame:
     """
@@ -153,11 +194,15 @@ def trim_dataframes(preprocessed: Dict[str, DataFrame], timerange,
     :return: Dict of trimmed dataframes
     """
     processed: Dict[str, DataFrame] = {}
-
     for pair, df in preprocessed.items():
-        trimed_df = trim_dataframe(df, timerange, startup_candles=startup_candles)
+        trimmed_df = trim_dataframe(df, timerange, startup_candles=startup_candles)
-        if not trimed_df.empty:
+        if not trimmed_df.empty:
-            processed[pair] = trimed_df
+            # start_mem = trimmed_df.memory_usage().sum() / 1024**2
+            # logger.info(f"Memory usage of df for {pair} before reduced is {start_mem:.2f} MB")
+            trimmed_df = reduce_mem_usage(pair, trimmed_df)
+            # end_mem = trimmed_df.memory_usage().sum() / 1024**2
+            # logger.info(f"Memory usage of df for {pair} after reduced is {end_mem:.2f} MB")
+            processed[pair] = trimmed_df
         else:
             logger.warning(f'{pair} has no data left after adjusting for startup candles, '
                            f'skipping.')