Init commit

ParaUtils/__init__.py

@@ -0,0 +1,60 @@
+#Bits and bobs used for many different projects. Junk drawer module
+
+#Useful functions for parsing race-like durations, and normalizing their string reps
+def t2f(S):
+    if not(S or False):
+        return None
+    m,s = ('0:%s' % S).split(':')[-2:]
+    try:
+        return np.round(60*int(m) + float(s),2)
+    except:
+        return -1
+
+def f2t(S):
+    if not(S or False) or (int(''.join([c for c in '0'+str(S) if c.isdigit()])) == 0) or (S <= 0):
+        return ''
+    S = float(S)
+    m = int(S/60)
+    s = S-60*m
+    hs = s-int(s)
+    s = int(s)
+    return f'{m:02}:{s:02}.{round(hs*100):02}'
+
+#Canada Summer Games eligible events
+CSG = pd.DataFrame([
+    {'Stroke':'Butterfly','Distance':50,'SportClass':[1,2,3,4,5,6,7]},
+    {'Stroke':'Butterfly','Distance':100,'SportClass':[8,9,10,11,12,13,14]},
+    {'Stroke':'Backstroke','Distance':50,'SportClass':[1,2,3,4,5]},
+    {'Stroke':'Backstroke','Distance':100,'SportClass':[1,2,6,7,8,9,10,11,12,13,14]},
+    {'Stroke':'Breaststroke','Distance':50,'SportClass':[1,2,3]},
+    {'Stroke':'Breaststroke','Distance':100,'SportClass':[4,5,6,7,8,9,11,12,13,14]},
+    {'Stroke':'Medley','Distance':150,'SportClass':[1,2,3,4]},
+    {'Stroke':'Medley','Distance':200,'SportClass':[5,6,7,8,9,10,11,12,13,14]},
+    {'Stroke':'Freestyle','Distance':50,'SportClass':[1,2,3,4,5,6,7,8,9,10,11,12,13,14]},
+    {'Stroke':'Freestyle','Distance':100,'SportClass':[1,2,3,4,5,6,7,8,9,10,11,12,13,14]},
+    {'Stroke':'Freestyle','Distance':200,'SportClass':[1,2,3,4,5,14]},
+    {'Stroke':'Freestyle','Distance':400,'SportClass':[6,7,8,9,10,11,12,13]},
+]).explode('SportClass').reset_index(drop=True)
+
+def parse_WPS_rankings(DF):
+    '''Takes a pandas datafram, usually loaded from an excel file, and cleans it up for use'''
+    return (
+        DF
+        #Makes .dot notation easier
+        .pipe(lambda df: df.rename(columns={c:c.replace(' ','_') for c in df.columns}))
+
+        #Drop VACANT events and relay results
+        .pipe(lambda df: df[df.SDMS_ID.notnull()])
+
+        #Parse out event information into useful columns
+        .pipe(lambda df: pd.concat([df,df.Event.str.extract(r"'s (?P<Distance>\d+) m (Individual )*(?P<Stroke>[A-z]+) [SBM]+(?P<SportClass>\d+)")],axis=1))
+
+        #Cleanup column dtypes
+        .assign(Date=lambda df: pd.to_datetime(df.Date))
+        .assign(Year=lambda df: df.Date.dt.year)
+        .assign(Seconds=lambda df: df.Result.apply(t2f))
+        .astype({'SportClass':int,'Distance':int,'SDMS_ID':int})
+
+        #Drop excess columns
+        .pipe(lambda df: df['Gender SportClass Distance Stroke Result Seconds SDMS_ID NPC Given_Name Family_Name Date'.split()])
+    )