okdolly-001

from datetime import datetime
startTime = datetime.now()

#do something

print datetime.now() - startTime

okdolly-001

#deleting a char directly in csv
infile = "test.csv"
outfile = "test_edit.csv"

delete_list = ["b'", "'"]
fin = open(infile)
fout = open(outfile, "w+")
for line in fin:
    for word in delete_list:
        line = line.replace(word, "")

okdolly-001

df.drop(df.std()[(df.std() == 0)].index, axis=1)

okdolly-001

# and save to csv
train = np.load('ecs171train.npy')
test = np.load('ecs171train.npy')
#Use atom and terminal to load the whole dataset if it's over 200+MB
train = train[:51000]
test = test[:51000]
np.savetxt('test.csv', test, fmt='%s')

okdolly-001

#Pandas deleting duplicate columns by value
df = pd.read_csv('data/test.csv')
df = df.loc[:, ~df.T.duplicated()]
df.to_csv('data/test.csv')

okdolly-001

clf = xgb.sklearn.XGBClassifier(
    objective="binary:logistic", 
    learning_rate=0.05, 
    seed=9616, 
    max_depth=20, 
    gamma=10, 
    n_estimators=500)

clf.fit(X_train, Y_train, early_stopping_rounds=20, eval_metric="auc", eval_set=eval_set, verbose=True)

okdolly-001

y_train['loss'] = (y_train['loss']  > 0.0).astype(int)

okdolly-001

#Prepend a line at the top of the file
def line_prepender(filename, line):
    with open(filename, 'r+') as f:
        content = f.read()
        f.seek(0, 0)
        f.write(line.rstrip('\r\n') + '\n' + content)


infile = "c_test.csv"
outfile = "complete_test.csv"

okdolly-001

1. Check if all elements in pandas/numpy are finite. notnull = isfinite

Pandas:
X_train.notnull().values.all()

np.isfinite(X_train).all()


2. Check if any elements in Pandas is Na.
Pandas:

okdolly-001

1. Substitute the NaN's in a dataframe with values from another dataframe

If you have two DataFrames of the same shape, then:

df[df.isnull()] = d2



2.Replace values in a dataframe with values from another dataframe by conditions