Ranked time series days with a total elo change metric

data_preparation/cleaned/time_series_days_ranked.csv

@@ -0,0 +1,18 @@
+,Date,DailyElo
+0,121,0.0
+1,122,-3.714599999999998
+2,178,0.04346000000000028
+3,179,2.1916710000000013
+4,180,0.0
+5,255,0.0
+6,256,0.0
+7,257,-2.520374784999996
+8,263,-2.0880156214999985
+9,264,-1.7032140593500005
+10,284,-0.6130256153877235
+11,285,2.620463284090865
+12,311,-2.076954630427971
+13,312,1.0960590427574828
+14,313,1.8954531384817344
+15,353,-0.2940921753664384
+16,354,-1.8646829578297937

elo_per_day.py

@@ -0,0 +1,33 @@
+import numpy as np
+import pandas as pd
+
+
+def join_cols():
+
+    # Reads in csv files to be manipulated
+    dfg = pd.read_csv('data_preparation/data/games_ranked.csv')
+
+    # Creates the new dataframe where each date is a unique column, and gets the number of dates
+    unique_dates = pd.DataFrame(dfg["Date"].unique()).to_numpy()
+    unique_rows = unique_dates.shape[0]
+    daily_elos = np.array(unique_rows).astype(float)
+    print(unique_rows)
+
+    # Creates two numpy arrays to perform some operations on
+    dates = dfg["Date"].to_numpy()
+    e_change = dfg["eloChangeAdjusted"].to_numpy()
+    rows = dates.shape()[0]
+
+    # sums up the elo change on a given day and then exports it to a unique .csv file
+    x = 0
+    for i in range(0, rows):
+        if not (dates[i] == unique_dates[x]):
+            x = x + 1
+        daily_elos[x] = daily_elos[x] + e_change[i]
+
+    # Creates a new dataframe from the two unique date array and the daily elo change array
+    df_dec = pd.DataFrame()
+    df_dec["Date"] = unique_dates
+    df_dec["DailyElo"] = daily_elos
+    print(df_dec)
+