diff --git a/src/main.py b/src/main.py index 1eb5b1a..1336b9d 100755 --- a/src/main.py +++ b/src/main.py @@ -90,7 +90,23 @@ def training(df): def clearData(df): res = df["class"].value_counts() + dtemp = df.sort_values(by=['class']) + supr = int(res["GALAXY"]/1.5) + + dtemp.drop(dtemp.index[range(1,supr)]) + dtemp = dtemp.iloc[34000:] + return dtemp + +def showDate(df): + res = df["class"].value_counts() + x = [res["GALAXY"],res["QSO"],res["STAR"]] + plt.figure(figsize = (8, 8)) + plt.pie(x, labels = ['GALAXY', 'QSO', 'Star']) + plt.legend() + + df=pd.read_csv('../data.csv') -clearData(df) +training(df) +