Skip to content

Instantly share code, notes, and snippets.

View mcguegi's full-sized avatar
🎯
Focusing

Camila Guerrero mcguegi

🎯
Focusing
View GitHub Profile
@mcguegi
mcguegi / metadata.csv
Created August 11, 2020 03:32
Ages and sexes of patients EEGs
edad sexo
46 f
66 f
55 m
41 f
63 f
72 f
31 m
25 f
42 f
@mcguegi
mcguegi / mean_data.csv
Created August 11, 2020 03:24
Mean of each frequency band per channel in exams EEGs
We can make this file beautiful and searchable if this error is corrected: It looks like row 6 should actually have 7 columns, instead of 5 in line 5.
channel,delta,tetha,alpha,beta,gamma,class
EEG O1-REFEEG O1-REFEEG O1-REFEEG O1-REFEEG O1-REFEEG O1-REFEEG O1-REFEEG O1-REFEEG O1-REFEEG O1-REF,0.1485798005359792,0.0900548946359901,0.2242448276011247,0.04434381846679583,0.018261265674641952,1
EEG F3-REFEEG F3-REFEEG F3-REFEEG F3-REFEEG F3-REFEEG F3-REFEEG F3-REFEEG F3-REFEEG F3-REFEEG F3-REF,0.100885884688964,0.03982581574915123,0.050565276457377385,0.022655879595769604,0.0195413421674164,1
EEG FZ-REFEEG FZ-REFEEG FZ-REFEEG FZ-REFEEG FZ-REFEEG FZ-REFEEG FZ-REFEEG FZ-REFEEG FZ-REFEEG FZ-REF,0.11813808348982324,0.030982591578498496,0.03604851875930125,0.017402092814197554,0.011459563533253784,1
EEG T4-REFEEG T4-REFEEG T4-REFEEG T4-REFEEG T4-REFEEG T4-REFEEG T4-REFEEG T4-REFEEG T4-REFEEG T4-REF,0.21108093314616247,0.10505401050935227,0.1371458510301313,0.048901857083469356,0.040938274229057756,1
EEG P4-REFEEG P4-REFEEG P4-REFEEG P4-REFEEG P4-REFEEG P4-REFEEG P4-REFEEG P4-REFEEG P4-REFEEG P4-REF,0.1340154450185936,0.07589951921127971,0.1582936805360596,0.0368957
@mcguegi
mcguegi / relative_power.csv
Created August 11, 2020 03:09
Relative band power dataset for EEGs clasiffication.
We can make this file beautiful and searchable if this error is corrected: It looks like row 9 should actually have 8 columns, instead of 1 in line 8.
exam,channel,class,alpha,betha,delta,gamma,theta
00002744_s002_t002.edf,EEG FZ-REF,0,0.00331264029234614,0.01927085562334667,0.9041643990974769,0.05925431676612765,0.00855319831585536
00002744_s002_t002.edf,EEG O2-REF,0,0.005708840055366444,0.023925826695482096,0.9014548418077574,0.054774694218668284,0.00881665916328943
00002744_s002_t002.edf,EEG FP2-REF,0,0.0059976608322190005,0.009865860069326708,0.9531339021926206,0.021581475594702314,0.008367924594705466
00002744_s002_t002.edf,EEG T6-REF,0,0.004286715802523175,0.018283142851878157,0.9188070379868494,0.047826831320528014,0.006073010485710719
00002744_s002_t002.edf,EEG O1-REF,0,0.009106224584520954,0.030342519319241982,0.878243881366932,0.06220264666278723,0.014379698929082127
00002744_s002_t002.edf,EEG F4-REF,0,0.00461364000054283,0.028009744977394505,0.8339344675953961,0.12138890703328395,0.0038606227780451567
00002744_s002_t002.edf,EEG P4-REF,0,0.003741669591018377,0.016004677873927092,0.9313113029166771,0.03940515374761982,0.006079478699094593
00002744_
@mcguegi
mcguegi / nn_loss_accuracy_graphs.py
Created July 1, 2020 22:11
Fragmento de código para graficar los epochs y loss-accuracy para redes neuronales
import matplotlib.pyplot as plt
import seaborn as sns
sns.set(font_scale=2)
fig, axes = plt.subplots(nrows=1, ncols=1, figsize=(30,10))
axes.plot(scores_train,color="#34495e")
axes.plot(scores_test,color="#2ecc71")
axes.set_title('Precisión del modelo vs Pasos temporales')
axes.set_ylabel('Valor')
axes.set_xlabel('Paso temporal')
@mcguegi
mcguegi / confusion_matrix_plots.py
Created July 1, 2020 21:00
Fragmento para las gráficas de matrices de confusión
import seaborn as sb
import matplotlib.pyplot as plt
import sklearn
sns.set(rc={'figure.figsize':(11.7,8.27)},style='whitegrid',font_scale=1.5,palette=sns.cubehelix_palette(8))
columnas = ['No epiléptico','Epiléptico']
df_cm = pd.DataFrame(confusion_matrix,index=columnas, columns=columnas)
grafica = sns.heatmap(df_cm,cmap=sns.cubehelix_palette(8),annot=True,fmt='g')
@mcguegi
mcguegi / signals_plots.py
Created July 1, 2020 03:55
Gráficos señales EEG (Dominio Tiempo/Frecuencia)
import matplotlib.pyplot as plt
import seaborn as sns
# COnfigurar tamaño de la letra y paleta de seaborn con la que se trabajará
sns.set(rc={'figure.figsize':(11.7,8.27)},style='whitegrid',font_scale=3,palette="PuBuGn_d")
# Figura con dos subgráficos
f, (ax1,ax2) = plt.subplots(2, 1, figsize=(30,10), sharex=True)
ax1.plot(channel_sano,c='#2b2d42')
ax1.set_title('Señal en el dominio del tiempo (Paciente No epiléptico)')