Calculate sample statistics. Changed names of CSV output files.

README.md

@@ -7,11 +7,11 @@ Data analysis tools for lifescale with GUI.
 The program *ls2csv* reads the content of the xlsm files written by lifescale units, parses the data and writes them to three csv 
 files (where `[run-name]` is the name from the settings sheet):
   * `Masses_Vibrio_[run-name].csv`: Data series from the sheet AcquisitionIntervals.
-  * `SampleMetadata_[run-name].csv`: Data from the sheet PanelData.
-  * `Summary_[run-name].csv`: Data from the sheet IntervalAnalysis plus sample related data from AcquisitionIntervals.
+  * `Metadata_[run-name].csv`: Data from the sheet PanelData.
+  * `SampleSummary_[run-name].csv`: Data from the sheet IntervalAnalysis plus sample related data from AcquisitionIntervals.
 
 ### Usage:
-  * Conversion: `ls2csv -i [path and nale of xlsm file] -o [outpur directory]`
+  * Conversion: `ls2csv -i [path and nale of xlsm file] -o [outpur directory] [-s] [-nv]`
   * Help: `ls2csv -h`
 
 

lifescale/command_line/command_line.py

@@ -36,12 +36,18 @@ def ls2csv():
     parser.add_argument("-o", "--out-dir", type=is_dir, required=True, help="Output directory for the CSV files.")
     parser.add_argument("-nv", "--not-verbose", required=False, help="Disable command line status messages.",
                         action='store_true')
+    parser.add_argument("-s", "--sample-stats", required=False, help="Calculate sample statistics (mean, median, IQR, "
+                                                                     "StDev.) and add it to the output CSV files.",
+                        action='store_true')
     # parser.add_argument("--out-dir", type=is_dir, required=False,
     #                     help="path to output directory", default=OUT_PATH)
     args = parser.parse_args()
     verbose = not args.not_verbose
 
-    return ls2csv_main(xlsm_filename=args.input_xlsm, output_dir=args.out_dir, verbose=verbose)
+    return ls2csv_main(xlsm_filename=args.input_xlsm,
+                       output_dir=args.out_dir,
+                       sample_stats=args.sample_stats,
+                       verbose=verbose)
 
 
 if __name__ == '__main__':

lifescale/models/ls_data.py

@@ -386,6 +386,8 @@ class LSData:
             zip(id_list, well_list, interval_num_list, time_list, masses_list, volume_list, total_num_particles_list,
                 transit_time_list, pressure_drop_list)),
             columns=df_masses_columns)
+        df_masses['Id'] = df_masses['Id'].astype(int)
+        df_masses['IntervalNumber'] = df_masses['IntervalNumber'].astype(int)
 
         # Sensor:
         # sensor_dict = {}
@@ -426,7 +428,7 @@ class LSData:
             filename_ending = f'{self.run_name}.csv'
 
         # Write PanelData:
-        filename = os.path.join(output_filepath, f'SampleMetadata_{filename_ending}')
+        filename = os.path.join(output_filepath, f'Metadata_{filename_ending}')
         if verbose:
             print(f'Write PanelData to: {filename}')
         self.df_panel_data.to_csv(filename, index=False)
@@ -444,7 +446,7 @@ class LSData:
             self.df_masses.to_csv(filename, index=False)
 
         # Write IntervalAnalysis:
-        filename = os.path.join(output_filepath, f'Summary_{filename_ending}')
+        filename = os.path.join(output_filepath, f'SamplesSummary_{filename_ending}')
         if verbose:
             print(f'Write IntervalAnalysis to: {filename}')
         self.df_interval_analysis.to_csv(filename, index=False)
@@ -453,6 +455,25 @@ class LSData:
         # TODO: Select columns for output (settable as parameter + default settings for each csv file)
         # TODO: Optionally order data series by time (parameter)?!
 
+    def calc_sample_statistics(self, verbose=True):
+        """Calculate statistical values for each sample and add it to the self.df_interval_analysis."""
+        if verbose:
+            print('Calculate sample statistics.')
+
+        for idx, row in self.df_interval_analysis.iterrows():
+            tmp_filter = (self.df_masses['Id'] == row.Id) & (self.df_masses['IntervalNumber'] == row.IntervalNumber)
+            tmp_filter_1 = (self.df_interval_analysis['Id'] == row.Id) & \
+                           (self.df_interval_analysis['IntervalNumber'] == row.IntervalNumber)
+            self.df_interval_analysis.loc[tmp_filter_1, 'Mass_median'] = self.df_masses.loc[tmp_filter, 'Mass'].median()
+            self.df_interval_analysis.loc[tmp_filter_1, 'Mass_std'] = self.df_masses.loc[tmp_filter, 'Mass'].std()
+            self.df_interval_analysis.loc[tmp_filter_1, 'Mass_q25'] = \
+                self.df_masses.loc[tmp_filter, 'Mass'].quantile(0.25)
+            self.df_interval_analysis.loc[tmp_filter_1, 'Mass_q75'] = \
+                self.df_masses.loc[tmp_filter, 'Mass'].quantile(0.75)
+            self.df_interval_analysis.loc[tmp_filter_1, 'Mass_iqr'] = \
+                self.df_interval_analysis.loc[tmp_filter_1, 'Mass_q75'] - \
+                self.df_interval_analysis.loc[tmp_filter_1, 'Mass_q25']
+
     @property
     def get_number_of_observations(self):
         """Return the number of observations (items in the data series)."""

lifescale/scripts/ls2csv.py

@@ -5,7 +5,9 @@ Copyright (C) 2022  Andreas Hellerschmied <heller182@gmx.at>"""
 from lifescale.models.ls_data import LSData
 
 
-def ls2csv(xlsm_filename, output_dir, verbose=True):
+def ls2csv(xlsm_filename, output_dir, sample_stats=True, verbose=True):
     """Convert lifescale output file (xlsm) to csv files."""
     ls_data = LSData.from_xlsm_file(input_xlsm_filename=xlsm_filename, verbose=verbose)
+    if sample_stats:
+        ls_data.calc_sample_statistics(verbose=verbose)
     ls_data.export_csv_files(output_dir, verbose=verbose)