def plot_histogram(same, diff):
df = {"PAINS vs. PAINS": same, "PAINS vs. ChEMBL": diff}
output_file("histogram_pains_self_diff.html")
hist = Histogram(df, bins=20, density=True, legend=True)
hist.x_range = Range1d(0, 1)
#hist.legend.orientation = "top_right"
show(hist)
def plot_actions_players_min(actions: dict, output_file_url: str) -> None:
"""
Plots all actions in a line diagrams and outputs them to an html file.
:param actions: a list of all actions and their timestamp.
:param output_file_url: string containing the directory and filename
:return: None
"""
# output to static HTML file
output_file(output_file_url)
tabs = []
for action in actions:
# create a new plot with a title and axis labels
p = Histogram(actions[action],
title=action,
bins=[0, 15, 30, 45, 60, 75, 90, 105])
p.xaxis.axis_label = 'Tijd (min.)'
p.yaxis.axis_label = 'Frequentie'
p.x_range = Range1d(0, 105)
# create a tab
tab = Panel(child=p, title=action)
# add to existing tabs
tabs.append(tab)
tabs = Tabs(tabs=tabs)
# show the results
show(tabs)
def compute_plot_all_Deff(tmin, tmax, particle_chemistry):
"""
Calculates and plots all Deffs in the timepoint range.
This function trims the cleaned MSD pandas dataframe to the user-
selected timepoint range, calculates a Deff from the mean MSD for
each timepoint within the range and for each particle chemistry,
plots a line graph of all Deffs across the timepoint range, and
gives geometric mean Deff values for each chemistry for the time
range specified.
Note: tmin must be less than tmax to get a timepoint range. If,
however, the user requires a Deff calculated from a single
timepoint, he or she can input an equal tmin and tmax, and the
function will consider only the single closest timepoint (on the
later side) to the input.
Inputs:
tmin: a float representing the minimum timepoint the user wishes to
consider in the Deff calculation
tmax: a float representing the maximum timepoint the user wishes to
consider in the Deff calculation (make this the same as tmin if a
single-timepoint Deff is desired)
particle_chemistry: a string matching the column title of the
particle chemistry which is to be emphasized: the Deffs of this
chemistry will be plotted on the histogram, bolded/dashed on the
line plot, and printed in the form of a geometric mean.
Outputs:
A bokeh histogram of Deffs for the inputted chemistry, on the same
figure as the below line plot (this figure defaults to fit nicely
in a 1280x800 screen, but plots can be resized as needed)
A bokeh line plot of Deffs for all particle chemistries across the
timepoint range, with emphasized chemistry bolded/dashed
Deff of inputted chemistry: a geometric mean of the Deffs between
tmin and tmax for the inputted chemistry, printed in the form of
a string within an explanatory sentence
avg_Deffs: a fresh pandas dataframe indexed with the columns
(particle chemistries) of the MSD dataframe, containing single
geometric mean Deff values for each particle chemistry
"""
# Verify time range validity
if tmin < 0 or tmax > round(max(msd.index)):
return "Error: input time range between 0 and your data's tmax"
else:
if tmin == 0:
print 'Divide by 0 error: tmin=0 changed to tmin=0.01'
tmin = 0.01
# Trim out-of-time-range rows
temp1_msd = msd[msd.index >= tmin]
# Trim to a length of 1 if user desires single-timepoint Deff
if tmin == tmax:
temp2_msd = temp1_msd.head(1)
else:
temp2_msd = temp1_msd[temp1_msd.index <= tmax]
# Calculate Deffs for only the timepoints needed and add as
# a new column to a new dataframe
index = temp2_msd.index
Deffs = pd.DataFrame(index=index, columns=columns2)
avg_Deffs = pd.DataFrame(index=columns2)
avg_Deffs_temp = []
h = Histogram([1], bins=50, width=300, height=250)
# maxes will eventually be used for sizing the histogram
maxes = []
# Lay the foundation for the bokeh figure
output_file('Deffs_hist_and_line_plot.html')
p = figure(
tools='resize,pan,box_zoom,wheel_zoom,reset,save', width=950,
height=650, x_axis_label='MSD timepoint', y_axis_label='Deff')
# Cycle through each particle chemistry and fill in Deffs
# and avg_Deffs
for title in columns2:
single_Deff_list = []
for i in range(0, len(temp2_msd)):
index = temp2_msd.index[i]
single_Deff_list.append(temp2_msd[title + ' geo'][index]/(
4*index**ALPHA))
# Add particle-chemistry-specific Deff list to Deffs
# dataframe
Deffs[title] = single_Deff_list
maxes.append(np.max(single_Deff_list))
# Add geometric mean Deff to what will become avg_Deffs
avg_Deffs_temp.append(scipy.stats.gmean(Deffs[title]))
# Create a special bold/dashed line for the inputted
# chemistry only, and generate the printed output
if title == particle_chemistry:
p.line(
Deffs.index, Deffs[title], line_width=5,
line_dash=(15, 10), legend=title, line_color=(
np.random.randint(256), np.random.randint(256),
np.random.randint(256))
)
h = Histogram(Deffs[particle_chemistry], width=300,
height=250)
print particle_chemistry + ' Deff = ' + str(
scipy.stats.gmean(Deffs[title]))
else:
p.line(
Deffs.index, Deffs[title], line_width=1, legend=title,
line_color=(
np.random.randint(256), np.random.randint(256),
np.random.randint(256))
)
avg_Deffs['Deff'] = avg_Deffs_temp
p.legend.label_text_font_size = '6pt'
# The line plot x-axis range is calibrated to include all
# of the desired data and the legend with no overlap
p.x_range = Range1d(tmin, tmax + (tmax - tmin)/5)
# The histogram x-axis is calibrated to remain at a
# constant scale for each given tmin/tmax combination--this
# gives a sense of scale for this particular histogram
# within the possible Deff values for all the chemistries
h.x_range = Range1d(0, np.max(maxes))
f = hplot(h, p)
show(f)
return avg_Deffs
def compute_hist_Deff(particle_chemistry, tmin, tmax):
"""
Calculates and plots Deff in timepoint range for a given chemistry.
This function trims the cleaned MSD pandas dataframe to the user-
selected timepoint range, calculates a Deff from the mean MSD for
each timepoint within the range, plots a histogram of the list of
Deffs, and gives a geometric mean Deff value for the time range
specified. This is all within the specified particle chemistry.
Note: tmin must be less than tmax.
Inputs:
particle_chemistry: a string matching the column title of the
particle chemistry for which a Deff is to be plotted and calculated
tmin: a float representing the minimum timepoint the user wishes to
consider in the Deff calculation
tmax: a float representing the maximum timepoint the user wishes to
consider in the Deff calculation
Outputs:
A bokeh charts histogram of Deff values calculated from the
particle chemistry's MSD of each timepoint within the range
Deff: a single float representing the geometric mean Deff value for
the timepoint range specified
Side effects:
The trimmed MSD dataset is appended (with a new column) to include
the list of Deffs for the given particle chemistry.
"""
# Verify time range validity
if tmin < 0 or tmax > round(max(msd.index)):
return "Error: input time range between 0 and your data's tmax"
else:
if tmin == 0:
print 'Divide by 0 error: tmin=0 changed to tmin=0.01'
tmin = 0.01
# Trim out-of-time-range rows
temp1_msd = msd[msd.index >= tmin]
temp2_msd = temp1_msd[temp1_msd.index <= tmax]
# Calculate Deffs for only the timepoints needed and add as a
# new column
Deff_list = []
for i in range(0, len(temp2_msd)):
index = temp2_msd.index[i]
# Calculate Deff using the conventional relationship
# between MSD and Deff
Deff_list.append(temp2_msd[particle_chemistry + ' geo'][index]/(
4*index**ALPHA))
temp2_msd[particle_chemistry + ' Deff'] = Deff_list
# Plot histogram and print mean Deff value
output_file('Deffs_hist.html')
p = Histogram(
temp2_msd[particle_chemistry + ' Deff'], bins=15, legend=False)
# Set range of x axis to reflect approximate range of all Deff
# values
p.x_range = Range1d(0, 0.015)
show(p)
# There's only one Deff column from this function: calculate
# geometric mean Deff from that column
Deff = scipy.stats.gmean(temp2_msd[particle_chemistry + ' Deff'])
return Deff
