feature add: Save pictures automatically when use ‘csikit -g’

CSIKit/tools/batch_graph.py

@@ -7,13 +7,16 @@
 from CSIKit.reader import get_reader, IWLBeamformReader
 
 DEFAULT_PATH = "./data/intel/misc/log.all_csi.6.7.6.dat"
+
+
 # DEFAULT_PATH = "./data/pi/walk_1597159475.pcap"
 
 class BatchGraph:
 
-    def __init__(self, path: str=DEFAULT_PATH, scaled: bool=False, filter_mac: str=None):
+    def __init__(self, path: str = DEFAULT_PATH, scaled: bool = False, filter_mac: str = None):
         reader = get_reader(path)
         self.csi_data = reader.read_file(path, scaled=scaled, filter_mac=filter_mac)
+        self.path = path
 
     def prepostfilter(self):
 
@@ -42,58 +45,48 @@ def prepostfilter(self):
         plt.ylabel("Amplitude (dBm)")
         plt.legend(loc="upper right")
 
-        plt.show()
+        output_path = self.path.replace('.dat', '_prepostfilter.png')
+        plt.savefig(output_path)
+        plt.close()
 
     def plotAllSubcarriers(self):
 
         finalEntry, no_frames, _ = get_CSI(self.csi_data)
 
         for x in finalEntry:
-            plt.plot(np.arange(no_frames)/20, x)
+            plt.plot(np.arange(no_frames) / 20, x)
 
         plt.xlabel("Time (s)")
         plt.ylabel("Amplitude (dBm)")
         plt.legend(loc="upper right")
 
-        plt.show()
+        output_path = self.path.replace('.dat', '_all_subcarriers.png')
+        plt.savefig(output_path)
+        plt.close()
 
     def heatmap(self):
 
-        # self.csi_data.frames = self.csi_data.frames[slice(1, len(self.csi_data.frames), 2)]
-
         finalEntry, no_frames, no_subcarriers = get_CSI(self.csi_data)
         if len(finalEntry.shape) == 4:
-            #>1 antenna stream.
-            #Loading the first for ease.
             finalEntry = finalEntry[:, :, 0, 0]
 
-        # from CSIKit.filters.wavelets.dwt import denoise
-        # finalEntry = denoise(finalEntry)
-
-        #Transpose to get subcarriers * amplitude.
         finalEntry = np.transpose(finalEntry)
 
         x_label = "Time (s)"
         try:
             x = self.csi_data.timestamps
-            x = [timestamp-x[0] for timestamp in x]
-        except AttributeError as e:
-            #No timestamp in frame. Likely an IWL entry.
-            #Will be moving timestamps to CSIData to account for this.
+            x = [timestamp - x[0] for timestamp in x]
+        except AttributeError:
             x = [0]
 
         if sum(x) == 0:
-            #Some files have invalid timestamp_low values which means we can't plot based on timestamps.
-            #Instead we'll just plot by frame count.
-
             xlim = no_frames
-
             x_label = "Frame No."
         else:
             xlim = max(x)
 
         limits = [0, xlim, 1, no_subcarriers]
-            
+
         _, ax = plt.subplots()
         im = ax.imshow(finalEntry, cmap="jet", extent=limits, aspect="auto")
 
@@ -105,34 +98,27 @@ def heatmap(self):
 
         plt.title(self.csi_data.filename)
 
-        plt.show()
+        output_path = self.path.replace('.dat', '_heatmap.png')
+        plt.savefig(output_path)
+        plt.close()
 
-    # Simple implementation.
-    # Takes a 2D matrix of CSI data and timestamps to visualise amplitude.
-    # CSI data should be arranged as (no_frames, no_subcarriers)
     @staticmethod
-    def plot_heatmap(csi_matrix, timestamps):
+    def plot_heatmap(csi_matrix, timestamps, output_path):
 
         csi_matrix = np.transpose(csi_matrix)
 
         x_label = "Time (s)"
         try:
             x = timestamps
             x = [timestamp - x[0] for timestamp in x]
-        except AttributeError as e:
-            # No timestamp in frame. Likely an IWL entry.
-            # Will be moving timestamps to CSIData to account for this.
+        except AttributeError:
             x = [0]
 
         if sum(x) == 0:
-            # Some files have invalid timestamp_low values which means we can't plot based on timestamps.
-            # Instead we'll just plot by frame count.
-
             xlim = csi_matrix.shape[1]
-
             x_label = "Frame No."
         else:
-            xlim = max(x)   
+            xlim = max(x)
 
         limits = [0, xlim, 1, csi_matrix.shape[0]]
 
@@ -147,40 +133,34 @@ def plot_heatmap(csi_matrix, timestamps):
 
         plt.title("CSI Amplitude Heatmap Plot")
 
-        plt.show()
+        plt.savefig(output_path)
+        plt.close()
 
     def sumsqrssi(self):
         finalEntry, no_frames, no_subcarriers = get_CSI(self.csi_data, extract_as_dBm=False)
         if len(finalEntry.shape) == 4:
-            # >1 antenna stream.
-            # Loading the first for ease.
             finalEntry = finalEntry[:, :, 0, 0]
 
-        rssis = []
-        sumsq = []
-
-
         csi = finalEntry
         rss = [x.rssi for x in self.csi_data.frames]
 
         sumsq = np.sum(csi ** 2, axis=1)
         norm_sumsq = np.sqrt(sumsq) / no_subcarriers
 
-        line = []
-
-        for rss_value, sumsq_value in zip(rss, norm_sumsq):
-            line.append(matlab.db(sumsq_value) / rss_value)
+        line = [matlab.db(sumsq_value) / rss_value for rss_value, sumsq_value in zip(rss, norm_sumsq)]
 
-        # plt.scatter(rssis, sumsq)
         plt.plot(self.csi_data.timestamps, line)
 
         plt.xlabel("Time (s)")
         plt.ylabel("RSSI / sumsq")
 
         plt.title(self.csi_data.filename)
 
-        plt.show()
+        output_path = self.path.replace('.dat', '_sumsqrssi.png')
+        plt.savefig(output_path)
+        plt.close()
+
 
 if __name__ == "__main__":
     bg = BatchGraph()
-    bg.heatmap()
+    bg.heatmap()