Quickstart: Parsing and Plotting

This example shows the most basic steps to get from a raw BioLector data file to an interactive visualization.

import pandas
import ipywidgets
from matplotlib import pyplot, cm

import bletl

In 98 % of the cases, you will use the bletl.parse function to load data.

The function comes with a few arguments that you can use to override calibration settings if your raw data file for example does not specify the correct lot_number.

To learn all about the different options, run help(bletl.parse) or bletl.parse?.

Parsing

In this quickstart, we load a raw data file from the test suite:

bldata = bletl.parse(r"..\..\..\tests\data\BLPro\213-NT_1500rpm_30C_20min--2019-10-18-14-55-15.csv")
# just display the variable containing the data
bldata

BLData(model=BLPro) {
  "BS4": FilterTimeSeries(197 cycles, 48 wells),
  "pH": FilterTimeSeries(197 cycles, 48 wells),
  "DO": FilterTimeSeries(197 cycles, 48 wells),
  "GFP-Gemini": FilterTimeSeries(197 cycles, 48 wells),
  "eYFP - Citrine": FilterTimeSeries(197 cycles, 48 wells),
}

As you can see in the above, the returned BLData object is a dictionary that maps identifiers of filtersets to FilterTimeSeries objects.

Metadata

Additional properties, such as comments, environment or filtersets contain data that is not specific to a well:

bldata.filtersets.head()

	filter_number	filter_name	filter_id	filter_type	excitation	emission	gain	gain_1	gain_2	phase_statistic_sigma	signal_quality_tolerance	reference_gain	reference_value	calibration	emission2
1	1	Biomass4	201	Intensity	None	None	NaN	4.0	1.0	None	None	NaN	NaN		None
2	2	pH(HP8)	202	pH	None	None	7.0	7.0	1.0	None	None	NaN	100.000000	1924_comp	None
3	3	DO(PSt3)	203	DO	None	None	7.0	7.0	1.0	None	None	NaN	100.000000	1924_comp	None
4	4	GFP-Gemini	204	Intensity	None	None	NaN	7.0	1.0	None	None	3.0	135.641129		None
5	5	eYFP - Citrine	215	Intensity	None	None	NaN	7.0	1.0	None	None	6.0	160.436661		None

bldata.environment.head()

	cycle	time	temp_setpoint	temp_up	temp_down	temp_water	o2	co2	humidity	shaker_setpoint	shaker_actual
5	1	0.013056	None	28.6	29.3	27.8	25.02	-9999.0	76.17	None	1500.0
156	1	0.076389	None	29.9	29.9	30.4	33.20	-9999.0	84.86	None	1500.0
207	1	0.095833	None	29.9	29.9	31.9	34.42	-9999.0	85.36	None	1499.0
285	2	0.346389	None	30.0	29.9	31.0	35.01	-9999.0	85.09	None	1499.0
436	2	0.409167	None	29.9	29.9	32.9	35.10	-9999.0	85.71	None	1500.0

Accessing and Visualizing Measurements

Depending on what you want to do, you can retrieve measurement data as tuples of Numpy arrays:

• bldata.get_timeseries

• bldata["pH"].get_timeseries

or as pandas DataFrames:

• bldata.get_narrow_data

• bldata.get_unified_narrow_data

• bldata["pH"].get_unified_dataframe

or via the underlying DataFrames that separate time and value:

• bldata["pH"].time

• bldata["pH"].value

# here we retrieve it as numpy arrays, becase that's very convenient for plotting
t, bs = bldata.get_timeseries("BS4", "A02")

pyplot.scatter(t, bs)
pyplot.xlabel('time   [h]')
pyplot.ylabel('backscatter 3')
pyplot.show()

Interactivity

With the ipywidgets package, one can easily create interactive visualizations.

Here, the plotting code from above is wrapped into a plot_one_well function that takes filterset and well as arguments.

The ipywidgets.interact or ipywidgets.interact_manual function is then used to create an interactive visualization, where all the available filtersets/wells are passed as options.

def plot_one_well(filterset, well):
    time, value = bldata.get_timeseries(filterset, well)

    fig, ax = pyplot.subplots()
    ax.scatter(time, value)
    ax.set_ylabel(filterset)
    ax.set_xlabel("time   [h]")
    ax.set_title(well)
    pyplot.show()
    return

ipywidgets.interact(
    plot_one_well,
    filterset=list(bldata.keys()),
    well=list(bldata["pH"].time.columns),
);

<function __main__.plot_one_well(filterset, well)>

That’s it with the basic tutorial! To learn more about what bletl can do, checkout the other examples.

Also remember to use help(...) and dir(...) to read the attributes/methods/docstrings. They often contain more detailed information than example notebooks.

%load_ext watermark
%watermark -n -u -v -iv -w

Last updated: Fri Jul 02 2021

Python implementation: CPython
Python version       : 3.7.9
IPython version      : 7.19.0

pandas    : 1.2.1
bletl     : 1.0.0
matplotlib: 3.3.2
ipywidgets: 7.6.3

Watermark: 2.1.0