|
| 1 | +# --- |
| 2 | +# jupyter: |
| 3 | +# jupytext: |
| 4 | +# formats: ipynb,py:percent |
| 5 | +# text_representation: |
| 6 | +# extension: .py |
| 7 | +# format_name: percent |
| 8 | +# format_version: '1.3' |
| 9 | +# jupytext_version: 1.18.1 |
| 10 | +# kernelspec: |
| 11 | +# display_name: Python 3 (ipykernel) |
| 12 | +# language: python |
| 13 | +# name: python3 |
| 14 | +# --- |
| 15 | + |
| 16 | +# %% [markdown] |
| 17 | +# # Title |
| 18 | +# |
| 19 | + |
| 20 | +# %% [markdown] |
| 21 | +# Description with link to ticket. |
| 22 | +# |
| 23 | + |
| 24 | +# %% |
| 25 | +# %load_ext autoreload |
| 26 | +# %autoreload 2 |
| 27 | +# %matplotlib inline |
| 28 | + |
| 29 | +# %% |
| 30 | +import logging |
| 31 | +import pathlib |
| 32 | + |
| 33 | +import dotenv |
| 34 | +import matplotlib |
| 35 | +import matplotlib.pyplot as plt |
| 36 | +import nest_asyncio |
| 37 | +import numpy as np |
| 38 | +import pandas as pd |
| 39 | +import seaborn as sns |
| 40 | +import ydata_profiling |
| 41 | +from IPython.core.interactiveshell import InteractiveShell |
| 42 | +from IPython.display import display, display_markdown, Markdown |
| 43 | +from markupspace import escape |
| 44 | +from pqdm.processes import pqdm |
| 45 | +# If you want threads instead: |
| 46 | +# from pqdm.threads import pqdm |
| 47 | +from tqdm.notebook import tqdm |
| 48 | + |
| 49 | +# %% |
| 50 | +# pd.options.mode.chained_assignment = None |
| 51 | +InteractiveShell.ast_node_interactivity = "all" |
| 52 | +# InteractiveShell.ast_node_interactivity = "last_expr" |
| 53 | +tqdm.pandas() |
| 54 | + |
| 55 | +# %% |
| 56 | + |
| 57 | +def printmd(string): |
| 58 | + display(Markdown(escape('# <span style="color:red">'+string+'</span>'))) |
| 59 | + |
| 60 | +display_markdown(escape('''## heading |
| 61 | +- ordered |
| 62 | +- list |
| 63 | +
|
| 64 | +The table below: |
| 65 | +
|
| 66 | +| id |value| |
| 67 | +|:---|----:| |
| 68 | +| a | 1 | |
| 69 | +| b | 2 | |
| 70 | +'''), raw=True) |
| 71 | + |
| 72 | +# %% [markdown] |
| 73 | +# Blue Alert Box: info. |
| 74 | +# <div class="alert alert-block alert-info"> |
| 75 | +# <b>Tip:</b> Use blue boxes (alert-info) for tips and notes. |
| 76 | +# If it's a note, you don't have to include the word "Note". |
| 77 | +# </div> |
| 78 | +# |
| 79 | +# # Yellow Alert Box: Warning. |
| 80 | +# <div class="alert alert-block alert-warning"> |
| 81 | +# <b>Example:</b> Yellow Boxes are generally used to include additional examples or mathematical formulas. |
| 82 | +# </div> |
| 83 | +# |
| 84 | +# # Green Alert Box: Success. |
| 85 | +# <div class="alert alert-block alert-success"> |
| 86 | +# Use green box only when necessary like to display links to related content. |
| 87 | +# </div> |
| 88 | +# |
| 89 | +# # Red Alert Box: Danger. |
| 90 | +# <div class="alert alert-block alert-danger"> |
| 91 | +# It is good to avoid red boxes but can be used to alert users to not delete some important part of code etc. |
| 92 | +# </div> |
| 93 | + |
| 94 | +# %% |
| 95 | +logger = logging.getLogger(__name__) |
| 96 | +logger.setLevel(logging.INFO) |
| 97 | +handler = logging.StreamHandler() |
| 98 | +# Add the handler to the logger |
| 99 | +logger.addHandler(handler) |
| 100 | + |
| 101 | +matplotlib.use('nbagg') |
| 102 | +# prettier plots |
| 103 | +plt.style.use('ggplot') |
| 104 | +# larger plots - two different ways. |
| 105 | +matplotlib.rc('figure', figsize=(15, 10)) |
| 106 | +plt.rcParams["figure.dpi"] = 90 |
| 107 | + |
| 108 | +# larger fonts |
| 109 | +sns.set_context('notebook', font_scale=1.5) |
| 110 | + |
| 111 | +dotenv.load_dotenv() |
| 112 | +nest_asyncio.apply() |
| 113 | + |
| 114 | +np.set_printoptions(precision=4) |
| 115 | +np.set_printoptions(suppress=True) |
| 116 | +pd.set_option('display.max_rows', 500) |
| 117 | +pd.options.display.max_columns = 50 |
| 118 | +pd.options.display.max_rows = 100 |
| 119 | +pd.options.display.max_colwidth = 80 |
| 120 | +# Adjust the number of columns profiled and displayed by the `info()` method. |
| 121 | +pd.options.display.max_info_columns = 150 |
| 122 | +# Adjust the number of decimals to be displayed in a DataFrame. |
| 123 | +pd.options.display.precision = 15 |
| 124 | +# Adjust the display format in a DataFrame. |
| 125 | +# pd.options.display.float_format = '{:.2f}%'.format |
| 126 | +# Prints and parses dates with the year first. |
| 127 | +pd.options.display.date_yearfirst = True |
| 128 | + |
| 129 | +InteractiveShell.ast_node_interactivity = "all" |
| 130 | + |
| 131 | +data_dir = pathlib.Path.cwd().parent / "data" |
| 132 | + |
| 133 | +def square(a): |
| 134 | + return a*a |
| 135 | + |
| 136 | +result = pqdm([1, 2, 3, 4, 5], square, n_jobs=2) |
| 137 | + |
| 138 | +profile = ydata_profiling.ProfileReport(df, title="Profiling Report") |
| 139 | +profile.to_notebook_iframe() |
| 140 | +# profile.to_widgets() |
| 141 | +# profile.to_html('profile.html') |
0 commit comments