Cheatography
https://cheatography.com
Pandas kudasaidzffffffsssssssssss
This is a draft cheat sheet. It is a work in progress and is not finished yet.
Import Data CSV and Excel
df = pd.read_csv('filename.csv') |
Read CSV into a Pandas DataFrame |
df = pd.to_csv('filename.csv') |
Export Pandas DataFrame to CSV |
df = pd.read_excel('filename.xlsx', sheet_name="Sheet 1") |
Read Excel into a Pandas DataFrame |
df = pd.to_excel('filename.xlsx', sheet_name="Sheet 1") |
Export Pandas DataFrame to Excel |
Import Options:
header=False, Index=False, usecols=(5,6)
Can also read CSV / HTML / JSON
Initial look into the DataFrame
df.head(5) |
Reads the first 5 rows |
df.tail(5) |
Reads the last 5 rows |
df.shape() |
Gives the number of columns and rows in the DataFrame |
df.dtypes |
Gives the datatypes for all the columns |
df['ColumnName'].dtypes |
Gives the datatypes for a single column |
df.head(-5) can retrieve last 5 lines similar to list[-5] retrieves the 5th last.
Change Column Data Type
df['col'] = df['col'].str.rstrip('%').astype('float') / 100.0 |
Remove % sign, convert to float and divide by 100 |
df['col'] = df['col'].str[:-1].astype('float') / 100.0 |
Blindly removing the last char - goes to the last character |
df['Column1']= df['Column1'].astype(float) |
Change 'Column1' to float |
|
|
Re-Order Colums
df = df[['Column3', 'Column2', 'Column1']] |
Re-orders the columns to the order specified in this list |
Dealing with NAN values
df= df.fillna(method='ffill') |
Fills blank values using forward fill method |
df= df.fillna(method='bfill') |
Fills blank values using backwards fill method |
df.dropna(inplace=True) |
Removes rows with no values |
Forward filling means fill missing values with last cell with value in the column. Backward filling means fill missing values with next cell with value.
Padding Values With Zero's
df['Column1'] = df['Column1'].astype(str).str.zfill(6) |
Sets the numbe to six (6) long, which adds zeros |
Obviously uses string since 0009 cannot be a valid number.
Filter Column
df2= df[(df['Column1'] >= some_number)] |
Filter DataFrame by certain value and putting it in another df. |
df2 = df[(df["Name"]=="Tom") & (df["Age"]==42)] |
Filter DataFrame by mulitple value and putting it in another df. |
|
