_allthingsdata_ @allthingsdata_

😁📈📊#dataanalytics #datarocks Berlin, Germany Joined November 2019

Tweets

31
Followers

57
Following

250
Likes

467

_allthingsdata_ @allthingsdata_

6 years ago

Python: Extract day from week number (YYYY-W01) format column monthly["new_col"] = monthly['weekn_column'].apply(lambda x: datetime.datetime.strptime(x + '-1', "%Y-W%W-%w")) Creates a new datetime column with YYYY-MM-DD.

0 0 0 0 0

_allthingsdata_ @allthingsdata_

6 years ago

0 0 0 0 0

Download Image

_allthingsdata_ @allthingsdata_

6 years ago

SQL lpad() function: LPAD('name', 7, 'sur'); Result: 'surname' Pads the left-side of a string with a specific set of characters

0 0 0 0 0

_allthingsdata_ @allthingsdata_

6 years ago

Python loop with zip(): def lzip(df): t = 0 for a, b in zip(df['A'], df['B']): t += a + b return t Takes several iterables making an iterator that aggragates elements from each iterable.

0 0 0 0 0

_allthingsdata_ @allthingsdata_

6 years ago

Python cut() function: pd.cut(df[‘col_name’], 6) Bins values into discrete intervals.

0 0 0 0 0

_allthingsdata_ @allthingsdata_

6 years ago

Python use of idxmin() and idxmax(): df.loc[df[‘Salary’].idxmax()][‘Name’]

0 0 0 0 0

_allthingsdata_ @allthingsdata_

6 years ago

SQL running totals using OVER clause: SELECT ID, Date, Amount, SUM(Amount) OVER(PARTITION BY Date ORDER BY ID) as RunningTotal FROM Sales ORDER BY ID, Amount

0 0 1 0 0

_allthingsdata_ @allthingsdata_

6 years ago

Python Pandas: Data Wrangling cheat sheet pandas.pydata.org/Pandas_Cheat_S…

0 0 0 0 0

_allthingsdata_ @allthingsdata_

6 years ago

Python counting occurrences: n = [1,3,2,7,5,2,3,3] get_threes = list(filter(lambda n: n == 3, n)) count_threes = len(threes)

0 0 0 0 0

_allthingsdata_ @allthingsdata_

6 years ago

Python sorting: my_array.sort() - sorts and mutates the array in place np.sort(my_array) - returns a copy of the sorted array without mutating the original array

0 0 1 0 0

_allthingsdata_ @allthingsdata_

6 years ago

Python most frequent: freq = [3, 5, 2, 2, 1, 2, 3, 2] print(max(set(freq), key = freq.count)) # 2 Finds the most frequently occurring value in a list or string.

0 0 2 0 0

_allthingsdata_ @allthingsdata_

6 years ago

SQL data loading in chunks: SELECT col_names FROM table ORDER BY col_1 OFFSET @PageSize * @PageNumber ROWS FETCH NEXT @PageSize ROWS ONLY; A paging query where @PageSize is the number of rows you want in each chunk and @PageNumber is the iteration number.

0 0 2 0 0

_allthingsdata_ @allthingsdata_

6 years ago

Python formatting: To simultaneously perform several independent replacements use - replace(). Eg. replace a comma and dollar sign with ' ' string, then convert column values to floats: df['col_name'] = df.col_name.replace('[$,]', ' ', regex = True).astype('float')