def observeStock(stockName, start_date, attribute='Close'):
'observe attribute of an arbitrary stock within the time specified'
'start_date and end_date must be in format %Y-%m-%d'
'attribute can be Open, High, Low, Close, Volume or Adj Close'
start_date = dt.strptime(start_date, "%d/%m/%Y").strftime("%Y-%m-%d")
end_date = curdate = (datetime.now() -
timedelta(days=1)).strftime("%Y/%m/%d")
sepwin = tkinter.Toplevel()
confr = tkinter.Frame(sepwin)
f = plt.figure(1, figsize=(10, 8))
df = web.DataReader(stockName, "yahoo", start_date, end_date)
X = df.index
Y = df[attribute]
plt.plot(X, Y)
plt.margins(0.02)
plt.xlabel('Date')
plt.ylabel(attribute)
canvas = FigureCanvasTkAgg(f, master=confr)
canvas.show()
canvas.get_tk_widget().pack(side=tkinter.TOP, fill=tkinter.BOTH, expand=1)
canvas.update_idletasks()
confr.pack()
class EnzymeCorrelatorGUI:
def __init__(self, root):
# Flags:
self.plot_only_lt = False
# GUI setup:
self.root = root
self.root.title('Enzyme Activity Correlator')
self.root.wm_attributes('-fullscreen', 1)
# Ininitialize frame style:
self.framestyle = ttk.Style()
self.framestyle.configure('TFrame', background='white')
# Create frames:
self.mainframe = ttk.Frame(self.root, padding=(0, 0, 12, 12))
# StringVars:
self.cutoff = tk.StringVar(self.mainframe, '0.85')
def update_cutoff(value):
self.sort_into_groups()
self.show_grouping_button_callback()
# Update histogram:
grouped_range = round((1 - float(self.cutoff.get())) / 0.05)
for i in range(len(self.patches)):
if i >= len(self.patches) - grouped_range:
self.patches[i].set_facecolor('indianred')
else:
self.patches[i].set_facecolor('steelblue')
self.canvas.draw()
# Mainframe widgets:
self.load_data_button = ttk.Button(self.mainframe,
text='Load Data',
command=self.load_data_callback)
self.show_grouping_button = ttk.Button(
self.mainframe,
text='Show Enzyme Grouping',
command=self.show_grouping_button_callback)
self.plot_correlation_matrix_button = ttk.Button(
self.mainframe,
text='Plot Correlation Matrix',
command=self.plot_correlation_data_callback)
self.plot_histogram_button = ttk.Button(
self.mainframe,
text='Plot Histogram',
command=self.plot_histogram_button_callback)
self.save_fig_button = ttk.Button(
self.mainframe,
text='Save Figure',
command=self.save_fig_button_callback)
self.quit_button = ttk.Button(self.mainframe,
text='Quit',
command=self.quit_button_callback)
self.grouping_label = tk.Text(root, height=10, width=150)
self.cutoff_slider = tk.Scale(root,
from_=-1,
to=1,
resolution=0.01,
variable=self.cutoff,
command=update_cutoff,
orient=tk.HORIZONTAL,
label='Set grouping cutoff')
# Mainframe grid management:
self.mainframe.grid(column=0, row=0, sticky=(tk.N, tk.S, tk.E, tk.W))
self.load_data_button.grid(column=0,
row=0,
sticky=(tk.N, tk.E, tk.W),
pady=5,
padx=5)
self.show_grouping_button.grid(column=0,
row=1,
sticky=(tk.N, tk.E, tk.W),
pady=5,
padx=5)
self.plot_correlation_matrix_button.grid(column=0,
row=2,
sticky=(tk.N, tk.E, tk.W),
pady=5,
padx=5)
self.plot_histogram_button.grid(column=0,
row=3,
sticky=(tk.N, tk.E, tk.W),
pady=(5, 0),
padx=5)
self.save_fig_button.grid(column=0,
row=4,
sticky=(tk.N, tk.E, tk.W),
pady=(5, 0),
padx=5)
self.quit_button.grid(column=0,
row=5,
sticky=(tk.N, tk.E, tk.W),
pady=(5, 0),
padx=5)
self.grouping_label.grid(column=1,
row=0,
columnspan=5,
sticky=(tk.N, tk.E, tk.W),
pady=5,
padx=5)
self.cutoff_slider.grid(column=1,
row=1,
columnspan=5,
sticky=(tk.N, tk.E, tk.W),
pady=5,
padx=5)
# Disable buttons until data has been loaded:
self.show_grouping_button["state"] = tk.DISABLED
self.plot_correlation_matrix_button["state"] = tk.DISABLED
self.plot_histogram_button["state"] = tk.DISABLED
self.save_fig_button["state"] = tk.DISABLED
self.cutoff_slider["state"] = tk.DISABLED
# Handle window resizing:
self.root.columnconfigure(0, weight=1)
self.root.rowconfigure(0, weight=0)
self.mainframe.columnconfigure(0, weight=3)
self.mainframe.rowconfigure(1, weight=0)
self.mainframe.rowconfigure(2, weight=0)
self.mainframe.rowconfigure(3, weight=0)
self.mainframe.rowconfigure(4, weight=0)
"""
DATA ANALYSIS
"""
def import_data(self):
enzyme_names = []
with open(self.datapath, mode='r') as csv_file:
csv_reader = csv.reader(csv_file, delimiter=';')
line_count = 0
for row in csv_reader:
if line_count == 0:
x_axis_labels = row[1:]
self.df = pd.DataFrame(data=None, columns=x_axis_labels)
line_count += 1
else:
row_data = []
enzyme = row.pop(0)
row_data.append(
[i.split(',')[0] + '.' + i.split(',')[1] for i in row])
enzyme_names.append(enzyme)
row_data = [float(i) for i in row_data[0]]
self.df = self.df.append(
pd.DataFrame([row_data],
columns=x_axis_labels,
dtype=float))
line_count += 1
self.df.index = enzyme_names
self.df = self.df.T
self.enzyme_list = []
for enzyme in self.df:
self.enzyme_list.append(self.df[enzyme])
def compute_correlation_matrix(self):
# Define columns for correlation matrix:
column_list = []
for enzyme in self.enzyme_list:
label = enzyme.name
column_list.append(label)
self.enzyme_matrix_columns = tuple(column_list)
# Calculate correlation matrix:
self.enzyme_correlation_matrix = np.zeros(
(len(self.enzyme_list), len(self.enzyme_list)))
for i in range(len(self.enzyme_matrix_columns)):
for j in range(len(self.enzyme_matrix_columns)):
if i >= j:
self.enzyme_correlation_matrix[i][j] = self.enzyme_list[
i].corr(self.enzyme_list[j]).round(decimals=2)
else:
if self.plot_only_lt is True:
self.enzyme_correlation_matrix[i][j] = 0
else:
self.enzyme_correlation_matrix[i][
j] = self.enzyme_list[i].corr(
self.enzyme_list[j]).round(decimals=2)
def compute_histogram(self):
self.hist_list = []
for i in range(len(self.enzyme_matrix_columns)):
for j in range(len(self.enzyme_matrix_columns)):
if i > j:
self.hist_list.append(self.enzyme_correlation_matrix[i][j])
binsize = 0.05
self.hist_axis = np.arange(-1, 1.05, binsize)
def sort_into_groups(self):
def corr_check(enzyme1, enzyme2):
if (enzyme1.corr(enzyme2) >= float(self.cutoff.get())
and enzyme1.name != enzyme2.name):
return True
else:
return False
# Set of all enzymes that have at least one correlation above cutoff with another enzyme:
correlating_enzymes = set()
for enzyme1 in self.enzyme_list:
for enzyme2 in self.enzyme_list:
if corr_check(enzyme1, enzyme2):
correlating_enzymes.add(enzyme1.name)
correlating_enzymes.add(enzyme2.name)
# Parse into mutually disjunct subsets:
grouped = {}
set_counter = 0
correlating_enzymes = list(correlating_enzymes)
for first in correlating_enzymes:
for new_partner in correlating_enzymes:
if corr_check(self.df[first], self.df[new_partner]):
# If one partner already belongs to a group -> carry over:
if (new_partner in grouped.keys()
and first not in grouped.keys()):
grouped[first] = grouped[new_partner]
elif (first in grouped.keys()
and new_partner not in grouped.keys()):
grouped[new_partner] = grouped[first]
# If both already belong to a different group -> merge:
elif (first in grouped.keys()
and new_partner in grouped.keys()):
# Transcribe all grouped[first]:
for enzyme in correlating_enzymes:
if enzyme in grouped.keys():
if grouped[enzyme] == grouped[first]:
grouped[enzyme] = grouped[new_partner]
# If none belong to any group -> create new:
else:
grouped[first] = set_counter
grouped[new_partner] = set_counter
set_counter += 1
# Put in correct order:
self.grouping = {}
groups = set(grouped.values())
for group_number in groups:
self.grouping[group_number] = []
for enzyme in correlating_enzymes:
self.grouping[grouped[enzyme]].append(enzyme)
# Correct for deletion of in-between group indeces:
counter = 0
for i in list(self.grouping.keys()):
self.grouping[counter] = self.grouping.pop(i)
counter += 1
"""
GUI CALLBACKS
"""
def load_data_callback(self):
self.datapath = tk.filedialog.askopenfilename(filetypes=(("csv files",
"*.csv"),
("all files",
"*.*")))
self.import_data()
self.compute_correlation_matrix()
self.compute_histogram()
self.sort_into_groups()
# Enable buttons:
self.show_grouping_button["state"] = tk.NORMAL
self.plot_correlation_matrix_button["state"] = tk.NORMAL
self.plot_histogram_button["state"] = tk.NORMAL
self.cutoff_slider["state"] = tk.NORMAL
def show_grouping_button_callback(self):
grouping_display = ''
grouping_display += "{:<8} {:<15}".format('Group', 'Enzymes')
for k, v in self.grouping.items():
a = str(v)
b = a[1:-1].replace("'", '')
grouping_display += '\n' + "{:<8} {:<15}".format(k, b)
# Clear before inserting new text to avoid overflow:
self.grouping_label.delete('1.0', tk.END)
self.grouping_label.insert(tk.END, grouping_display)
def plot_correlation_data_callback(self):
self.fig = Figure(figsize=(19, 9))
ax = self.fig.add_subplot(111)
im = ax.imshow(self.enzyme_correlation_matrix,
aspect='auto',
cmap='bwr')
im.set_clim(-1, 1)
ax.grid(False)
if self.plot_only_lt is True:
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
ax.xaxis.set(ticks=tuple(np.arange(0, len(self.enzyme_list), 1)),
ticklabels=self.enzyme_matrix_columns)
ax.tick_params(axis="x", rotation=45, labelsize=9)
ax.yaxis.set(ticks=tuple(np.arange(0, len(self.enzyme_list), 1)),
ticklabels=self.enzyme_matrix_columns)
ax.set_ylim(len(self.enzyme_list) - 0.5, -0.5)
for i in range(len(self.enzyme_list)):
for j in range(len(self.enzyme_list)):
if i < j:
if self.plot_only_lt is True:
color = 'white'
else:
color = 'black'
else:
color = 'black'
ax.text(j,
i,
self.enzyme_correlation_matrix[i][j],
ha='center',
va='center',
color=color,
size=9)
ax.figure.colorbar(im, ax=ax, format='% .2f')
try:
self.canvas.get_tk_widget().grid_forget()
except AttributeError:
pass
self.canvas = FigureCanvasTkAgg(self.fig, master=self.root)
self.canvas.get_tk_widget().grid(columnspan=2,
sticky=(tk.N, tk.E, tk.W),
pady=5,
padx=5)
self.canvas.draw()
# Enable save_fig_button:
self.save_fig_button["state"] = tk.NORMAL
def plot_histogram_button_callback(self):
self.fig = Figure(figsize=(20, 5))
ax = self.fig.add_subplot(111)
N, bins, self.patches = ax.hist(self.hist_list,
bins=self.hist_axis,
color='steelblue',
ec='k')
grouped_range = round((1 - float(self.cutoff.get())) / 0.05)
for i in range(len(self.patches) - grouped_range, len(
self.patches)): # This should be inferred from self.cutoff
self.patches[i].set_facecolor('indianred')
ax.set_xticks(self.hist_axis[::2])
# plt.yticks(np.arange(0, 25, 2))
ax.grid(True)
ax.set_xlim([-1, 1])
ax.grid(color='black', linestyle=':', linewidth=0.25)
ax.set_xlabel('Correlation of activity between enzyme pairs')
ax.set_ylabel('Occurrence')
try:
self.canvas.get_tk_widget().grid_forget()
self.canvas.update_idletasks()
except AttributeError:
pass
self.canvas = FigureCanvasTkAgg(self.fig, master=self.root)
self.canvas.get_tk_widget().grid(columnspan=2,
sticky=(tk.N, tk.E, tk.W),
pady=5,
padx=5)
self.canvas.draw()
# Enable save_fig_button:
self.save_fig_button["state"] = tk.NORMAL
def save_fig_button_callback(self):
savename = tk.filedialog.asksaveasfilename()
self.fig.savefig(savename)
pass
def quit_button_callback(self):
self.root.destroy()