�� : ��
��

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from google.colab import drive
drive.mount('/content/drive')
Mounted at /content/drive
df = pd.read_csv('drive/MyDrive/iris_extended.csv')
df.sample(10)
	species	elevation	soil_type	sepal_length	sepal_width	petal_length	petal_width	sepal_area	petal_area	sepal_aspect_ratio	...	sepal_to_petal_length_ratio	sepal_to_petal_width_ratio	sepal_petal_length_diff	sepal_petal_width_diff	petal_curvature_mm	petal_texture_trichomes_per_mm2	leaf_area_cm2	sepal_area_sqrt	petal_area_sqrt	area_ratios
931	virginica	251.5	loamy	6.70	3.33	4.63	1.94	22.3110	8.9822	2.012012	...	1.447084	1.716495	2.07	1.39	7.96	10.80	56.34	4.723452	2.997032	2.483913
132	setosa	150.2	clay	5.18	3.11	1.56	0.33	16.1098	0.5148	1.665595	...	3.320513	9.424242	3.62	2.78	5.37	22.10	59.84	4.013702	0.717496	31.293318
353	setosa	178.8	clay	4.71	4.20	1.53	0.58	19.7820	0.8874	1.121429	...	3.078431	7.241379	3.18	3.62	5.05	22.61	57.08	4.447696	0.942019	22.292089
376	setosa	203.9	clay	5.14	4.13	1.52	0.31	21.2282	0.4712	1.244552	...	3.381579	13.322581	3.62	3.82	5.49	18.79	51.04	4.607407	0.686440	45.051358
329	setosa	285.2	sandy	4.68	2.95	1.23	0.20	13.8060	0.2460	1.586441	...	3.804878	14.750000	3.45	2.75	4.45	19.58	46.68	3.715643	0.495984	56.121951
154	setosa	144.7	clay	5.43	3.83	1.59	0.38	20.7969	0.6042	1.417755	...	3.415094	10.078947	3.84	3.45	4.58	18.94	54.00	4.560362	0.777303	34.420556
957	virginica	162.2	loamy	6.34	2.84	6.15	2.22	18.0056	13.6530	2.232394	...	1.030894	1.279279	0.19	0.62	13.44	10.58	72.82	4.243301	3.694997	1.318802
355	setosa	190.5	loamy	4.71	3.09	1.42	0.34	14.5539	0.4828	1.524272	...	3.316901	9.088235	3.29	2.75	4.30	22.60	40.25	3.814957	0.694838	30.144780
263	setosa	256.1	loamy	4.76	3.17	1.62	0.37	15.0892	0.5994	1.501577	...	2.938272	8.567568	3.14	2.80	5.30	19.91	58.38	3.884482	0.774209	25.173841
303	setosa	111.7	sandy	5.17	3.33	1.50	0.36	17.2161	0.5400	1.552553	...	3.446667	9.250000	3.67	2.97	5.22	19.83	57.71	4.149229	0.734847	31.881667

df2 = df

for col in df2.columns:
  if df2[col].dtype == float:
    q1 = df2[col].quantile(0.25)
    q3 = df2[col].quantile(0.75)
    iqr = q3 - q1

    up = q3 + 1.5 * iqr
    down = q1 - 1.5 * iqr
    print(col, up, down)

    df2 = df2[(df2[col] <= up) & (df2[col] >= down)]

print('ejection_count', len(df.index) - len(df2.index))

all = len(df2.index)

# clay sandy loamy
# virginica setosa versicolor

counts = df2.species.value_counts()

all_clay = df2[df2.soil_type == 'clay'].species.value_counts()
all_sandy = df2[df2.soil_type == 'sandy'].species.value_counts()
all_loamy = df2[df2.soil_type == 'loamy'].species.value_counts()

virginica_counts = [all_clay['virginica'] / counts['virginica'] * 100,
                    all_sandy['virginica'] / counts['virginica'] * 100,
                    all_loamy['virginica'] / counts['virginica'] * 100]

setosa_counts = [all_clay['setosa'] / counts['setosa'] * 100,
                    all_sandy['setosa'] / counts['setosa'] * 100,
                    all_loamy['setosa'] / counts['setosa'] * 100]

versicolor_counts = [all_clay['versicolor'] / counts['versicolor'] * 100,
                    all_sandy['versicolor'] / counts['versicolor'] * 100,
                    all_loamy['versicolor'] / counts['versicolor'] * 100]

barWidth = 0.25
fig = plt.subplots(figsize =(12, 8))

br1 = np.arange(3)
br2 = [x + barWidth for x in br1]
br3 = [x + barWidth for x in br2]

plt.bar(br1, virginica_counts, color ='black', width = barWidth,
        edgecolor ='white', label ='���������')
plt.bar(br2, setosa_counts, color ='orange', width = barWidth,
        edgecolor ='white', label ='������')
plt.bar(br3, versicolor_counts, color ='grey', width = barWidth,
        edgecolor ='white', label ='����������')

plt.xlabel('����� �����������', fontsize = 15)
plt.ylabel('����� �����, ��������', fontsize = 15)
plt.xticks([r + barWidth for r in range(3)],
        ['��������� �����', '�������� �����', '����������� �����'])

plt.legend()
plt.show()
df2.sample(20)

y = pd.Series(list(df2.petal_texture_trichomes_per_mm2))
x = pd.Series(list(df2.leaf_area_cm2))
correlation = y.corr(x)

plt.scatter(x, y, s=1)

plt.plot(np.unique(x), np.poly1d(np.polyfit(x, y, 1))
         (np.unique(x)), color='red')
plt.scatter(x, y, s=1, c='b')
plt.plot(np.unique(x),
         np.poly1d(np.polyfit(x, y, 1))
         (np.unique(x)), color='red')
plt.xlabel('������� ����� � ���������� �����������')
plt.ylabel('���-�� �������� �� ���������� ���������')
Text(0, 0.5, '���-�� �������� �� ���������� ���������')
��
�� .csv
�� .docx
��
��������� ���� ��������

��
