%load_ext google.colab.data_table import google import numpy as np import pandas as pd import matplotlib.pyplot as plt import seaborn as sns import altair as alt from IPython.display import display from pprint import pprint

!pip install japanize-matplotlib > /dev/null import japanize_matplotlib

from google.cloud import bigquery google.colab.auth.authenticate_user() project_id = "" client = bigquery.Client(project=project_id) query = ''' ''' df = client.query(query).to_dataframe()

# テーブル名から table = bigquery.TableReference.from_string(bq_table) bqclient.list_rows(table).to_dataframe(create_bqstorage_client=True)

import google google.colab.auth.authenticate_user() creds, _ = google.auth.default() import gspread sheet_url = '' gs = gspread.authorize(creds) worksheet = gs.open_by_url(sheet_url).get_worksheet(0) # worksheet = gs.open_by_url(...).worksheet('Sheet1') df = pd.DataFrame(worksheet.get_all_values()) ''' # 前にデータ以外がある & 末尾に合計がある時の整形 cell = worksheet.find("ヘッダ左上セル") start_row = cell.row columns = worksheet.row_values(start_row) columns df = pd.DataFrame(worksheet.get_all_values()[start_row:-1], columns=[]) '''

pd.read_csv('https://docs.google.com/spreadsheets/d/{ID}/export?format=csv')

import google creds, _ = google.auth.default() # requests from google.auth.transport.requests import AuthorizedSession authed_session = AuthorizedSession(credentials) response = authed_session.get('...') # colab の urlib3 は古くてサンプルコードはそのまま動かない !pip install -q -U google-auth urllib3 import urllib3 from google.auth.transport.urllib3 import AuthorizedHttp authed_http = AuthorizedHttp(creds)

import tensorflow as tf import tensorflow_probability as tfp

import IPython app = IPython.Application.instance() app.kernel.do_shutdown(True)

job_config = bigquery.QueryJobConfig() job_config.dry_run = True job = client.query(query, job_config) # 以下のように処理バイト数見れる fields = ['total_bytes_processed', 'total_bytes_billed'] for field in fields: print(field, getattr(job, field))

job_config = bigquery.QueryJobConfig() job_config.maximum_bytes_billed = 5_000_000_000 # 5GB result = client.query(query, job_config) # <- Error when it scans over than 5GB

from google.colab import data_table data_table.DataTable(data.airports(), include_index=False, num_rows_per_page=10)

from google.colab import files uploaded = files.upload()

uploaded[filename] # bytes list(uploaded.values())[0] # ファイル名を入力するのダルいので1ファイルならこう読むかな import pandas as pd import io bytes = list(uploaded.values())[0] data = pd.read_csv(io.StringIO(bytes.decode('utf-8'))) data.head()

from google.colab import drive drive.mount('/drive') df.to_csv('/drive/My Drive/test.csv') # pickle !mkdir -p '/drive/My Drive/colab/20230402_hoge/' df.to_pickle('/drive/My Drive/colab/20230402_hoge/df.pkl')

with open(f'/content/output.tsv', 'w', encoding="utf-8") as f: f.write(tsv)

text = 'value' # @param {type:"string"} dropdown = '1st option' #@param ["1st option", "2nd option", "3rd option"] date_input = '2018-03-22' #@param {type:"date"} boolean_checkbox = True #@param {type:"boolean"} boolean_dropdown = True #@param ["False", "True"] {type:"raw"}

import itables itables.show(df) # 折り返しするには css でやる... css = """ <style> td { white-space: normal !important; word-wrap: break-word !important; } </style> """ # CSSを適用 from IPython.core.display import HTML display(HTML(css))

import google from google.auth.transport.requests import AuthorizedSession google.colab.auth.authenticate_user() creds, project = google.auth.default() client = AuthorizedSession(creds) resp = client.get(f'https://www.googleapis.com/drive/v2/files/{doc_id}/comments')

from IPython.display import HTML, display def print_monospaced(text: str): html_output = f'<pre style="font-family: monospace;">{text}</pre>' display(HTML(html_output))