import numpy as np               # linear algebra
import pandas as pd              # data processing, CSV file I/O (e.g. pd.read_csv)
from sklearn.metrics import pairwise_distances
from joblib import Parallel, delayed
import seaborn as sns
import plotly as plt

input_file = "tiny.csv"
output_file = f"results_jaccard_{input_file}"
passeport = "passport_fonio.txt"

## passeport fonio

pass_df= pd.read_csv(passeport, sep=' ')

pass_df.tail()

	ID	country	species
242	Q27_02_02	Nigeria	Digitaria exilis
243	Q27_04_01	Nigeria	Digitaria exilis
244	Q31_02_02	Nigeria	Digitaria exilis
245	Q31_04_01	Nigeria	Digitaria exilis
246	Q32_07_02	Nigeria	Digitaria exilis

pass_df['ID'] = pass_df['ID'].str.replace('_','')
pass_df['ID'] = pass_df['ID'].str.replace('-','')
pass_df['ID'] = pass_df['ID'].str.replace('NER2011','NER')
pass_df['ID'] = pass_df['ID'].str.replace('MNHNPP','PP')

pass_df.tail()

	ID	country	species
242	Q270202	Nigeria	Digitaria exilis
243	Q270401	Nigeria	Digitaria exilis
244	Q310202	Nigeria	Digitaria exilis
245	Q310401	Nigeria	Digitaria exilis
246	Q320702	Nigeria	Digitaria exilis

pass_df.groupby(['species']).count()

	ID	country
species
Digitaria exilis	199	199
Digitaria iburua	21	21
Digitaria longiflora	13	13
Digitaria ternata	11	11
other	2	2

grouped = pass_df.groupby(['species'])

for name, group in grouped:
    print(name)
    print(len(group["ID"].tolist()))

('Digitaria exilis',)
199
('Digitaria iburua',)
21
('Digitaria longiflora',)
13
('Digitaria ternata',)
11
('other',)
2

results_file = "SPLITTED/jaccard_100x10000.txt"

import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
import statsmodels.formula.api as smf
import statsmodels.api as sm

df_s = pd.read_csv(results_file, sep="\t", decimal='.')
df_s['Pop'] = df_s['Pop'].str.replace('Digitaria','D ')
df_s.head()

	input_file	Kmers_nb	Pop	Size_pop	Jaccard_Average_Disimilarity
0	xaa.txt	50000	D exilis	199	0.305329
1	xaa.txt	50000	D exilis	199	0.303862
2	xaa.txt	50000	D exilis	199	0.299224
3	xaa.txt	50000	D exilis	199	0.297954
4	xaa.txt	50000	D exilis	199	0.299823

On veut étudier ici l'influence de la pop de fonio sur la disimilarity de jaccard.

sns.set()
ax = sns.boxplot(x="Pop", y="Jaccard_Average_Disimilarity", data=df_s, color='blue')
plt.xlabel('Pop')
plt.ylabel('Jaccard_Disimilarity')
#plt.title('Boîtes à moustaches')
plt.show()

gp = df_s.groupby(['Pop'])
with open('tmp_stats_pop_fonio.txt', "w") as fd:
    fd.write(f'Population\tJacc_mean\tJaccard_var\tJaccard_var_corr\tJaccard_std\tJaccard_coeffvariation\tMean-2*std\tMean+2*std\n')
    for name, group in gp:
        fd.write(
            f"{name[0]}\t{group.Jaccard_Average_Disimilarity.mean()}\t{group.Jaccard_Average_Disimilarity.var()}\t{group.Jaccard_Average_Disimilarity.var(ddof=1)}\t{group.Jaccard_Average_Disimilarity.std()}\t{(group.Jaccard_Average_Disimilarity.std()/group.Jaccard_Average_Disimilarity.mean())}\t{group.Jaccard_Average_Disimilarity.mean() - 2*(group.Jaccard_Average_Disimilarity.std())}\t{group.Jaccard_Average_Disimilarity.mean() + 2*(group.Jaccard_Average_Disimilarity.std())}\n")

summary_stats = pd.read_csv('tmp_stats_pop_fonio.txt', sep="\t", decimal='.')
summary_stats

	Population	Jacc_mean	Jaccard_var	Jaccard_var_corr	Jaccard_std	Jaccard_coeffvariation	Mean-2*std	Mean+2*std
0	D exilis	0.283280	0.000628	0.000628	0.025053	0.088440	0.233174	0.333386
1	D iburua	0.431045	0.001138	0.001138	0.033740	0.078274	0.363566	0.498524
2	D longiflora	0.378836	0.000355	0.000355	0.018839	0.049730	0.341157	0.416515
3	D ternata	0.439251	0.002042	0.002042	0.045185	0.102869	0.348881	0.529622
4	other	0.938903	0.000150	0.000150	0.012257	0.013055	0.914388	0.963418

df_s.boxplot(column="Jaccard_Average_Disimilarity", by="Pop", vert=False)
plt.show()

sns.violinplot(data=df_s, x="Jaccard_Average_Disimilarity", y="Pop")

<Axes: xlabel='Jaccard_Average_Disimilarity', ylabel='Pop'>