Size Distribution Stats¶
This example shows how to process size distribution data from an SMPS. The processing returns mean properties of the size distribution, such as the mean diameter, median diameter, and total PM2.5 mass.
# all the imports
import matplotlib.pyplot as plt
from particula_beta.data import loader_interface, settings_generator
from particula_beta.data.tests.example_data.get_example_data import get_data_folder
from particula_beta.data.loader_setting_builders import (
# These functions create settings for loading data from files.
DataChecksBuilder,
SizerDataReaderBuilder,
LoaderSizerSettingsBuilder,
)
# the new step
from particula_beta.data.process import size_distribution
# set the parent directory of the data folder
path = get_data_folder()
Load the data¶
For this example we'll use the provided example data. But you can change the path to any folder on your computer. We then can used the settings generator to
If you think this settings generator is getting tedious, we hear you. We'll show the fix to that soon.
# settings for the SMPS data
data_checks_sizer = (
DataChecksBuilder()
.set_characters([250])
.set_skip_rows(25)
.set_char_counts({"/": 2, ":": 2})
.build()
)
data_sizer_reader = (
SizerDataReaderBuilder()
.set_sizer_start_keyword("20.72")
.set_sizer_end_keyword("784.39")
.set_sizer_concentration_convert_from("dw/dlogdp")
.build()
)
smps_1d_settings, smps_2d_settings = (
LoaderSizerSettingsBuilder()
.set_relative_data_folder("SMPS_data")
.set_filename_regex("*.csv")
.set_header_row(24)
.set_data_checks(data_checks_sizer)
.set_data_column(
[
"Lower Size (nm)",
"Upper Size (nm)",
"Sample Temp (C)",
"Sample Pressure (kPa)",
"Relative Humidity (%)",
"Median (nm)",
"Mean (nm)",
"Geo. Mean (nm)",
"Mode (nm)",
"Geo. Std. Dev.",
"Total Conc. (#/cm³)",
]
)
.set_data_header(
[
"Lower_Size_(nm)",
"Upper_Size_(nm)",
"Sample_Temp_(C)",
"Sample_Pressure_(kPa)",
"Relative_Humidity_(%)",
"Median_(nm)",
"Mean_(nm)",
"Geo_Mean_(nm)",
"Mode_(nm)",
"Geo_Std_Dev.",
"Total_Conc_(#/cc)",
]
)
.set_data_sizer_reader(data_sizer_reader)
.set_time_column([1, 2])
.set_time_format("%m/%d/%Y %H:%M:%S")
.set_delimiter(",")
.set_timezone_identifier("UTC")
.build()
)
# collect settings into a dictionary
combined_settings = {
'smps_1d': smps_1d_settings,
'smps_2d': smps_2d_settings,
}
# now call the loader interface for files
lake = loader_interface.load_folders_interface(
path=path,
folder_settings=combined_settings,
)
print(' ')
print(lake)
Folder Settings: smps_1d Loading file: 2022-07-07_095151_SMPS.csv Loading file: 2022-07-10_094659_SMPS.csv Folder Settings: smps_2d Loading file: 2022-07-07_095151_SMPS.csv Loading file: 2022-07-10_094659_SMPS.csv Lake with streams: ['smps_1d', 'smps_2d']
Processing the Stream¶
The lake is a collection of streams, stored as a dictionary. The next step
can be average the data or you may have a processing step that you want to
apply to all the data. For example, you may want to calculate the total
PM2.5 mass from the SMPS data. You could do this by looping through the
streams and applying a cusutom processing function to each stream. Or you could
use some standard process already built in to particula_beta.data.process. In this
example we'll use process.size_distribution to calculate the PM2.5 mass from the
lake['mean_properties'] = size_distribution.sizer_mean_properties(
stream=lake['smps_2d'],
diameter_units='nm',
)
# list out the header
for header in lake['mean_properties'].header:
print(header)
Total_Conc_(#/cc) Mean_Diameter_(nm) Geometric_Mean_Diameter_(nm) Mode_Diameter_(nm) Mean_Diameter_Vol_(nm) Mode_Diameter_Vol_(nm) Unit_Mass_(ug/m3) Mass_(ug/m3) Total_Conc_(#/cc)_N100 Unit_Mass_(ug/m3)_N100 Mass_(ug/m3)_N100 Total_Conc_(#/cc)_PM1 Unit_Mass_(ug/m3)_PM1 Mass_(ug/m3)_PM1 Total_Conc_(#/cc)_PM2.5 Unit_Mass_(ug/m3)_PM2.5 Mass_(ug/m3)_PM2.5 Total_Conc_(#/cc)_PM10 Unit_Mass_(ug/m3)_PM10 Mass_(ug/m3)_PM10
Plot the Data¶
With that processing done we can plot some useful summary plots. For example, we can plot the total PM2.5 mass as a function of time. And on the same plot we can add the N100 mass.
Tip: Calling the header directly¶
Note below how we can call the data header directly from the stream. This is because the get items property defined in the stream class accepts, either and index or a header name. So we can call the header name directly from the stream and get back that specific time series.
This is incontrast to callint stream.data['header_name'] which would return
an error. As that line first calls stream.data returning the np.ndarray, then calls
the header name, which is not a valid index for a np.ndarray.
mean_prop_stream = lake['mean_properties']
# plot the data on twinx axis
fig, ax = plt.subplots()
ax.plot(mean_prop_stream.datetime64,
mean_prop_stream['Mass_(ug/m3)_PM2.5'],
label='PM 2.5',
color='blue')
ax.plot(mean_prop_stream.datetime64,
mean_prop_stream['Mass_(ug/m3)_N100'],
label='N100 mass',
color='red')
ax.set_ylim(0, 50)
plt.xticks(rotation=45)
ax.set_xlabel("Time (UTC)")
ax.set_ylabel('PM mass (ug/m3)')
ax.legend()
plt.show()
fig.tight_layout()
Summary¶
This example showed how to process size distribution data from an SMPS. The processing returns mean properties of the size distribution, such as the mean diameter, median diameter, and total PM2.5 mass.