quantile-scatter 0.3.4

Visualization tool that makes it easier to get scatter plots right.

quantile_scatter

下の方に日本語の説明があります

Overview

• Visualization tool that makes it easier to get scatter plots right.
• The number of uniform data is divided into intervals on the x-axis, and the quantile points for each interval are displayed.

Usage

import quantile_scatter

# dummy data
x_ls = [(4 * random.random() - 2) ** 3
	for _ in range(1000)]
y_ls = [math.sin(x) + random.random() * 0.5
	for x in x_ls]

# plot [quantile_scatter]
quantile_scatter.plot(
	x = x_ls,	# x-list
	y = y_ls,	# y-list
	min_bin_ratio = 1/20,	# Ratio of the smallest group (the number of records in the smallest group as a percentage of the total)
	ile_ls = [0.25, 0.5, 0.75]
)

Advanced Usage

• Option argument of quantile_scatter.plot() function:

mean = True   # Also draw the "mean"
show = False  # Do not show the graph and only return the data to be displayed (useful for saving the graph or drawing with something other than matplotlib)
missing_values = [None]	# The specification that x contains a missing value of "None". Internally, the value is included in the statistics as "missing".

概要

• 散布図を正しく把握しやすくする可視化ツール
• 均一データ数の横軸区間に分け、各区間の分位点を表示する
• 説明は執筆中です

使用例

import quantile_scatter

# ダミーデータ
x_ls = [(4 * random.random() - 2) ** 3
	for _ in range(1000)]
y_ls = [math.sin(x) + random.random() * 0.5
	for x in x_ls]

# 分位点散布図の描画 [quantile_scatter]
quantile_scatter.plot(
	x = x_ls,	# 横軸数値リスト
	y = y_ls,	# 縦軸数値リスト
	min_bin_ratio = 1/20,	# 最小グループ割合 (最も小さいグループのレコード数が全体に占める割合)
	ile_ls = [0.25, 0.5, 0.75]	# どこの分位点を出すか
)

発展的な利用方法

quantile_scatter.plot()関数のoption引数

mean = True	# 「平均」も描画する
show = False	# グラフ表示せず、表示対象データのみを返却 (グラフを保存したい場合や、matplotlib以外で描画したい場合などに有効)
missing_values = [None]	# xにNoneという欠損値が含まれるという指定。内部的には"missing"という値として集計に含められる。