Python Pandas CSV Data numpy


16
1
df_full = pd.read_csv('./datasets/housing_train.csv')
2

3

4
df_full = df_full.select_dtypes(include=[np.number]).dropna()
5

6
df_y = df_full.iloc[:,'SalePrice']
7

8
df_y = df_full.ix[:,'SalePrice']
9
df_x = df_full.ix[:,['YearBuilt', 'OverallQual’]]
10

11
df_x = df_full.ix[:100, ‘Col’]
12
df_x = df_full.ix[:100] # whole rows
13

14
df_y = df_full.loc['SalePrice']
15

16
np_dataset_x = df_x.values

Filter rows (negation)


1
1
df_threshold = df_threshold[~((df_threshold.ground_truth == 'neutral') & (df_threshold.confidence_pred == 'neutral|noise'))]

Access rows


15
1
df.loc[df['column_name'] == some_value]
2

3
To select rows whose column value is in an iterable, some_values, use isin:
4

5
df.loc[df['column_name'].isin(some_values)]
6
Combine multiple conditions with &:
7

8
df.loc[(df['column_name'] >= A) & (df['column_name'] <= B)]
9
Note the parentheses. Due to Python's operator precedence rules, & binds more tightly than <= and >=. Thus, the parentheses in the last example are necessary. Without the parentheses
10

11
df['column_name'] >= A & df['column_name'] <= B
12
is parsed as
13

14
df['column_name'] >= (A & df['column_name']) <= B
15

Apply


1
1
df["Grades"].apply(lambda val: "Yes" if val < 45 else "No")

Priekš Data science ļoti useful


8
1
pd.options.display.max_columns = 999
2
pd.options.display.max_rows = 999
3

4
print(df["classes"].value_counts())
5
# .to_dict()
6

7
print(df_full.describe())
8
print(df["selling_price"].describe())