Merge pull request #16 from meyravDayan/master

.

.vscode/settings.json

@@ -0,0 +1,3 @@
+{
+    "python.pythonPath": "C:\\ProgramData\\Miniconda3\\envs\\my_env\\python.exe"
+}

submovements/DataProcessing.py

@@ -6,6 +6,87 @@
 import numpy as np
 from scipy.signal import butter, filtfilt
 from pathlib import Path
+import re
+import seaborn as sns;
+
+sns.set()
+
+
+@attr.s
+class Subject(object):
+    dir_path = attr.ib(default='')
+    id = attr.ib(init=False)
+    df_folder = attr.ib(default='')
+    df_dict = attr.ib({'square_left': [], 'square_right': [], 'tri_left': [], 'tri_right': []})
+
+    def __attrs_post_init__(self):
+        self.id = os.path.split(self.dir_path)[1]
+
+    def create_total_df(self):
+        self.df_total = pd.DataFrame({'Vx': [], 'Vy': [], 'Rep': [], 'Block': [], 'Time': [], 'Condition': [],
+                                      'ID': [], 'pos x': [],
+                                      'pos y': []})  # creating an empty array for concatination use later
+        pproc = Preprocessor()
+        trial_gen = pproc.load_df_from_directory_gen(self.dir_path)
+        for trial in trial_gen:
+            trial.preprocess(pproc)
+            df = trial.create_df()
+            trial.save_df(df, self.df_folder)
+            for key in self.df_dict:
+                if key == trial.stimulus:
+                    self.df_dict[key].append(df)
+            self.df_total = pd.concat([self.df_total, df])
+        self.df_total = self.df_total.set_index(['ID', 'Condition', 'Block', 'Rep', 'Time']).sort_values(
+            ['ID', 'Condition', 'Block', 'Rep'], ascending=True)
+
+    def stimuli_plot_vx(self, num_of_trials):
+        fig, axes = plt.subplots(2, 2)  # add name to subplots
+        plt.xlabel('Time (sec)')
+        plt.ylabel('Vx (cm/sec)')
+        for n, key in enumerate(self.df_dict):
+            if n == 0:
+                axes[0, 0].set_title(key)
+            if n == 1:
+                axes[0, 1].set_title(key)
+            if n == 2:
+                axes[1, 0].set_title(key)
+            if n == 3:
+                axes[1, 1].set_title(key)
+            for trial in range(num_of_trials):
+                if n == 0:
+                    sns.lineplot(x="Time", y="Vx", data=self.df_dict[key][trial], ax=axes[0, 0])
+                if n == 1:
+                    sns.lineplot(x="Time", y="Vx", data=self.df_dict[key][trial], ax=axes[0, 1])
+                if n == 2:
+                    sns.lineplot(x="Time", y="Vx", data=self.df_dict[key][trial], ax=axes[1, 0])
+                if n == 3:
+                    sns.lineplot(x="Time", y="Vx", data=self.df_dict[key][trial], ax=axes[1, 1])
+        plt.show()
+
+    def stimuli_plot_vy(self, num_of_trials):
+        fig, axes = plt.subplots(2, 2)  # add name to subplots
+        plt.xlabel('Time (sec)')
+        plt.ylabel('Vy (cm/sec)')
+        for n, key in enumerate(self.df_dict):
+            if n == 0:
+                axes[0, 0].set_title(key)
+            if n == 1:
+                axes[0, 1].set_title(key)
+            if n == 2:
+                axes[1, 0].set_title(key)
+            if n == 3:
+                axes[1, 1].set_title(key)
+            for trial in range(num_of_trials):
+                if n == 0:
+                    sns.lineplot(x="Time", y="Vy", data=self.df_dict[key][trial], ax=axes[0, 0])
+                if n == 1:
+                    sns.lineplot(x="Time", y="Vy", data=self.df_dict[key][trial], ax=axes[0, 1])
+                if n == 2:
+                    sns.lineplot(x="Time", y="Vy", data=self.df_dict[key][trial], ax=axes[1, 0])
+                if n == 3:
+                    sns.lineplot(x="Time", y="Vy", data=self.df_dict[key][trial], ax=axes[1, 1])
+        plt.show()
+
 
 @attr.s
 class Trial():
@@ -58,20 +139,20 @@ def create_df(self):
         # Vx, Vy, Condition, Time, ID, Block, Repetition
         # TODO: Generalize to z axis
         ###
-
+
+        pos_x = self.data['x']
+        pos_y = self.data['y']
         vx = self.data['Vx']
         vy = self.data['Vy']
-        time = np.arange(len(vx))    # change later !!!!
+        time = self.data.index.values
         condition = np.full(shape=len(vx), fill_value=self.stimulus)
-
-        # maybe add to main as trial attribute
         id = np.full(shape=len(vx),
                      fill_value=self.id
                      )
         block = np.full(shape=len(vx), fill_value=self.block, dtype=np.int)
         rep = np.full(shape=len(vx), fill_value=self.rep, dtype=np.int)
         return pd.DataFrame({'Vx': vx, 'Vy': vy, 'Rep': rep, 'Block': block,
-                             'Time': time, 'Condition': condition, 'ID': id})
+                             'Time': time, 'Condition': condition, 'ID': id, 'pos x': pos_x, 'pos y': pos_y})
 
     def save_df(self, df, dest_folder):
         ###
@@ -98,7 +179,7 @@ class Preprocessor():
     # cut-via-threshold and plot DataFrames
     ###
 
-    raw_paths = attr.ib(default='data') # Path list of all the raw data files
+    raw_paths = attr.ib(default='data')  # Path list of all the raw data files
     sample_rate = attr.ib(default=240)  # Sample rate the raw data is sampled at
 
     raw_headers = attr.ib(default=['SampleNum', 'x', 'y', 'z',
@@ -145,7 +226,7 @@ def load_single_file(self, file_path, cols=('x', 'y')):
 
         assert Path(file_path).is_file(), f'File does not exists: {file_path}'
         file_path = Path(file_path)
-        cols=list(cols)
+        cols = list(cols)
         trial_out = Trial()
         try:
             df = pd.read_csv(file_path, names=self.raw_headers)
@@ -245,11 +326,12 @@ def remove_baseline(self, data_in: pd.DataFrame, cols=('x', 'y'), threshold=0.05
         idx = df.loc[df >= threshold]
 
         # expand data cutting limits
-        low_cut_index = int(idx.index[0]-0.1*self.sample_rate \
-            if df.index.min() < idx.index[0]-0.1*self.sample_rate \
-            else df.index.min())
-        high_cut_index = int(idx.index[-1]+0.1*self.sample_rate \
-            if df.index.max() > idx.index[-1]+0.1*self.sample_rate \
-            else df.index.max())
+
+        low_cut_index = int(idx.index[0] - 0.1 * self.sample_rate \
+                                if df.index.min() < idx.index[0] - 0.1 * self.sample_rate \
+                                else df.index.min())
+        high_cut_index = int(idx.index[-1] + 0.1 * self.sample_rate \
+                                 if df.index.max() > idx.index[-1] + 0.1 * self.sample_rate \
+                                 else df.index.max())
 
         return data_in.copy()[low_cut_index:high_cut_index].reset_index(drop=True)

submovements/meyrav_work.py

@@ -0,0 +1,15 @@
+from DataProcessing import Preprocessor, Trial, Subject
+import pandas as pd
+import numpy as np
+import os
+import re
+import attr
+import matplotlib.pyplot as plt
+from scipy.signal import savgol_filter, hilbert
+import seaborn as sns; sns.set()
+
+if __name__ == "__main__":
+    a = Subject('../1', r'C:\Users\meyra\Desktop\s')
+    a.create_total_df()
+    a.stimuli_plot_vx(10)
+    a.stimuli_plot_vy(10)