datafest competition 2019
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 

80 lines
2.7 KiB

import pandas as pd
def vectorize_mult(df, column, dictionary):
"""
Changes all the categorical values into its respective
number in the dictionary and then saves it in the DF
:param df: dataframe
:param column: column name
:param dictionary: alterations to make
"""
newCol = column + "Num"
df[newCol] = df[column].map(dictionary)
class WellnessCSV:
def __init__(self):
self.file = "data/wellness.csv"
self.end = "cleaned/notnormalized_with_0NaN_wellness.csv"
def vectorize(self):
df = pd.read_csv(self.file)
# Vectorizing appropriate data
vectorize_mult(df, "Pain", {"No": 0, "Yes": 1})
vectorize_mult(df, "Illness", {"No": 0, "Slightly Off": 0.5, "Yes": 1})
vectorize_mult(df, "Menstruation", {"No": 0, "Yes": 1})
vectorize_mult(df, "Nutrition", {"Poor": 0, "Okay": 0.5, "Excellent": 1})
vectorize_mult(df, "NutritionAdjustment", {"No": 0, "Yes": 1})
vectorize_mult(df, "USGMeasurement", {"No": 0, "Yes": 1})
readiness = []
for i, value in df["TrainingReadiness"].iteritems():
value = value.split("%")[0]
value = int(value) * (1/100)
readiness.append(value)
df["TrainingReadinessNum"] = readiness
# Filling in NaNs for appropriate layers where they won't make a statistical difference
df["MenstruationNum"] = df["MenstruationNum"].fillna(0)
df["USGMeasurementNum"] = df["USGMeasurementNum"].fillna(0)
df["NutritionNum"] = df["MenstruationNum"].fillna(0)
df["NutritionAdjustmentNum"] = df["NutritionAdjustmentNum"].fillna(0)
# Saving the df to the "cleaned" CSV
df.to_csv(self.end)
class FatigueSum:
def __init__(self):
self.file = "cleaned/time_series_normalized_wellness.csv"
self.end = "cleaned/fatigue_total_sum.csv"
def calculate(self):
df = pd.read_csv(self.file)
# get some of the fatigue for a particular date
diction = dict()
dates = df["TimeSinceAugFirst"].unique()
dates = set(dates)
dates = list(dates)
# for each date, get unique data and get calculation
for date in dates:
pdf = df[df["TimeSinceAugFirst"] == date]
num_players = len(pdf["playerID"].unique())
fatigue_sum = pdf["normFatigue"].sum()
result = fatigue_sum / num_players
diction[date] = result
# Converting
dates = diction.keys()
values = diction.values()
final_df = pd.DataFrame()
final_df["TimeSinceAugFirst"] = dates
final_df["fatigueSum"] = values
final_df.to_csv(self.end)