Added initial analysis output.

app.py

@@ -36,10 +36,12 @@
 import datetime
 import deepfakeecg
 import ecg_plot
+import getopt
 import gradio
 import io
 import matplotlib.pyplot as plt
 import matplotlib.ticker
+import neurokit2
 import numpy
 import pathlib
 import random
@@ -242,8 +244,34 @@ def analyze(request: gradio.Request):
    log(f'Session "{request.session_hash}": Analyze ECG #{Sessions[request.session_hash].Selected + 1}!')
 
    data = Sessions[request.session_hash].Results[Sessions[request.session_hash].Selected]
+
+   data = data.t().detach().cpu().numpy()[1:] / 1000
+
+   print(data[0])
+   print(len(data[0]))
+
+   leadI = data[0]
+
+   signals, info = neurokit2.ecg_process(leadI, sampling_rate = deepfakeecg.ECG_SAMPLING_RATE)
+   neurokit2.ecg_plot(signals, info)
+
+   # DIN A4 landscape: w=11.7, h=8.27
+   # w = 508/25.4   # mm to inch
+   # h = 122/25.4   # mm to inch
+   # matplotlib.pyplot.gcf().set_size_inches(w, h, forward=True)
+   # pdf.savefig(matplotlib.pyplot.gcf())
+   # matplotlib.pyplot.close()
+
    print(data)
 
+   return matplotlib.pyplot.gcf()
+
+
+# ###### Print usage and exit ###############################################
+def usage(exitCode = 0):
+   sys.stdout.write('Usage: ' + sys.argv[0] + ' [-d|--device cpu|cuda] [-v|--version]\n')
+   sys.exit(exitCode)
+
 
 
 # ###### Main program #######################################################
@@ -312,6 +340,36 @@ img.logo-image {
 """
 
 
+# ====== Check arguments ====================================================
+try:
+   options, args = getopt.getopt(
+      sys.argv[1:],
+      'd:v',
+      [
+         'device=',
+         'version'
+      ])
+   for option, optarg in options:
+      print(option)
+      if option in ( '-d', '--device' ):
+         runOnDevice = optarg
+      elif option in ( '-v', '--version' ):
+         sys.stdout.write('PyTorch version: ' + torch.__version__ + '\n')
+         sys.stdout.write('CUDA version:    ' + torch.version.cuda + '\n')
+         sys.stdout.write('CUDA available:  ' + ('yes' if torch.cuda.is_available() else 'no') + '\n')
+         sys.stdout.write('Device:          ' + runOnDevice + '\n')
+         sys.exit(1)
+      else:
+         sys.stderr.write('ERROR: Invalid option ' + option + '!\n')
+         sys.exit(1)
+
+except getopt.GetoptError as error:
+   sys.stderr.write('ERROR: ' + str(error) + '\n')
+   usage(1)
+if len(args) > 0:
+   usage(1)
+
+
 # ====== Create GUI =========================================================
 with gradio.Blocks(css = css, theme = gradio.themes.Glass(secondary_hue=gradio.themes.colors.blue)) as gui:
 
@@ -356,6 +414,8 @@ with gradio.Blocks(css = css, theme = gradio.themes.Glass(secondary_hue=gradio.t
                                      preview    = True)
       outputGallery.select(select)
    gradio.Markdown('## Analysis')
+   with gradio.Row():
+      analysisOutput = gradio.Plot()
 
    # ====== Add click event handling for "Generate" button ==================
    buttonGenerate.click(predict,
@@ -367,19 +427,22 @@ with gradio.Blocks(css = css, theme = gradio.themes.Glass(secondary_hue=gradio.t
                        )
 
    # ====== Add click event handling for "Analyze" button ===================
-   buttonAnalyze.click(analyze)
+   buttonAnalyze.click(analyze,
+                       inputs = [],
+                       outputs = [ analysisOutput ]
+                      )
 
    # ====== Add click event handling for download buttons ===================
    # Using hidden button and JavaScript, to generate download file on-the-fly:
    # https://github.com/gradio-app/gradio/issues/9230#issuecomment-2323771634
    buttonCSV.click(downloadCSV)
    buttonCSV.click(fn = downloadCSV, inputs = None, outputs = [ buttonCSV_hidden ]).then(
-                      fn = None, inputs = None, outputs = None,
-                      js = "() => document.querySelector('#download_csv_hidden').click()")
+                   fn = None, inputs = None, outputs = None,
+                   js = "() => document.querySelector('#download_csv_hidden').click()")
    buttonPDF.click(downloadPDF)
    buttonPDF.click(fn = downloadPDF, inputs = None, outputs = [ buttonPDF_hidden ]).then(
-                      fn = None, inputs = None, outputs = None,
-                      js = "() => document.querySelector('#download_pdf_hidden').click()")
+                   fn = None, inputs = None, outputs = None,
+                   js = "() => document.querySelector('#download_pdf_hidden').click()")
 
    # ====== Run on startup ==================================================
    gui.load(predict,