Skip to content

Commit

Permalink
updates to plotting to use user defined attributes
Browse files Browse the repository at this point in the history
  • Loading branch information
@carolinesands
carolinesands committed Nov 13, 2023
1 parent 9362b6a commit 7c774c9
Show file tree
Hide file tree
Showing 9 changed files with 286 additions and 249 deletions.
28 changes: 27 additions & 1 deletion nPYc/StudyDesigns/SOP/Generic.json
View file
Original file line number Diff line number Diff line change
Expand Up @@ -10,5 +10,31 @@
"featureMetadataNotExported":[],
"analyticalMeasurements":{},
"excludeFromPlotting":[],
"sampleTypeColours":{"StudySample": "b", "StudyPool": "g", "ExternalReference": "r", "MethodReference": "m", "ProceduralBlank": "c", "Other": "grey", "Study Sample": "b", "Study Reference": "g", "Long-Term Reference": "r", "Method Reference": "m", "Blank": "c", "Unspecified SampleType or AssayRole": "grey"}
"sampleTypeColours":{
"StudySample": "blue", "Assay": "blue", "Study Sample": "blue", "Sample": "blue",
"StudyPool": "darkgreen", "Study Reference": "darkgreen",
"ExternalReference": "darkorange", "Long-Term Reference": "darkorange", "Long Term Reference": "darkorange",
"LinearityReference": "red", "Linearity Reference": "red",
"MethodReference": "blue", "Method Reference": "blue",
"ProceduralBlank": "aquamarine", "Blank": "aquamarine",
"UnknownType": "grey", "UnknownRole": "grey", "Unknown": "grey", "nan": "grey", "NaN": "grey", "NA": "grey"
},
"sampleTypeMarkers":{
"StudySample": "o", "Assay": "o", "Study Sample": "o", "Sample": "o",
"StudyPool": "v", "Study Reference": "v",
"ExternalReference": "^", "Long-Term Reference": "^", "Long Term Reference": "^",
"LinearityReference": "s", "Linearity Reference": "s",
"MethodReference": "P", "Method Reference": "P",
"ProceduralBlank": "X", "Blank": "X",
"UnknownType": "d", "UnknownRole": "d", "Unknown": "d", "nan": "d", "NaN": "d", "NA": "d"
},
"sampleTypeAbbr":{
"StudySample": "SS", "Assay": "SS", "Study Sample": "SS", "Sample": "SS",
"StudyPool": "SR", "Study Reference": "SR",
"ExternalReference": "LTR", "Long-Term Reference": "LTR", "Long Term Reference": "LTR",
"LinearityReference": "SRD", "Linearity Reference": "SRD",
"MethodReference": "MR", "Method Reference": "MR",
"ProceduralBlank": "Blank", "Blank": "Blank",
"UnknownType": "NA", "UnknownRole": "NA", "Unknown": "NA", "nan": "NA", "NaN": "NA", "NA": "NA"
}
}
16 changes: 16 additions & 0 deletions nPYc/objects/_msDataset.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1330,6 +1330,22 @@ def _getSampleMetadataFromFilename(self, filenameSpec):
fileNameParts.loc[
fileNameParts['groupingKind'].str.match('E?IC', na=False).astype(bool), 'SampleType'] = SampleType.StudyPool

# Standardised NPC types based on SampleType/AssayRole combinations
fileNameParts.loc[(fileNameParts['SampleType'] == SampleType.StudySample) & (
fileNameParts['AssayRole'] == AssayRole.Assay), 'SampleClass'] = 'Study Sample'
fileNameParts.loc[(fileNameParts['SampleType'] == SampleType.StudyPool) & (
fileNameParts['AssayRole'] == AssayRole.PrecisionReference), 'SampleClass'] = 'Study Reference'
fileNameParts.loc[(fileNameParts['SampleType'] == SampleType.ExternalReference) & (
fileNameParts['AssayRole'] == AssayRole.PrecisionReference), 'SampleClass'] = 'Long-Term Reference'
fileNameParts.loc[(fileNameParts['SampleType'] == SampleType.StudyPool) & (
fileNameParts['AssayRole'] == AssayRole.LinearityReference), 'SampleClass'] = 'Linearity Reference'
fileNameParts.loc[(fileNameParts['SampleType'] == SampleType.MethodReference) & (
fileNameParts['AssayRole'] == AssayRole.PrecisionReference), 'SampleClass'] = 'Method Reference'
fileNameParts.loc[(fileNameParts['SampleType'] == SampleType.ProceduralBlank) & (
fileNameParts['AssayRole'] == AssayRole.Blank), 'SampleClass'] = 'Blank'
fileNameParts.loc[(fileNameParts['SampleType'] == SampleType.UnknownType) & (
fileNameParts['AssayRole'] == AssayRole.UnknownRole), 'SampleClass'] = 'Unknown'

# Skipped runs
fileNameParts['Skipped'] = fileNameParts['exclusion'].str.match('[Xx]', na=False)

Expand Down
120 changes: 50 additions & 70 deletions nPYc/plotting/_multivariatePlotting.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -59,13 +59,13 @@ def plotScree(R2, Q2=None, title = '', xlabel='', ylabel='', savePath=None, figu
else:
plt.show()

def plotScores(pcaModel, classes=None, classType=None, components=None, alpha = 0.05, plotAssociation=None, title ='', xlabel='', figures=None, savePath=None, figureFormat='png', dpi=72, figureSize=(11, 7)):
def plotScores(pcaModel, classes=None, colourType=None, colourDict=None, components=None, alpha = 0.05, plotAssociation=None, title ='', xlabel='', figures=None, savePath=None, figureFormat='png', dpi=72, figureSize=(11, 7)):
"""
Plot PCA scores for each pair of components in PCAmodel, coloured by values defined in classes, and with Hotelling's T2 ellipse (95%)
:param ChemometricsPCA pcaModel: PCA model object (scikit-learn based)
:param pandas.Series classes: Measurement/groupings associated with each sample, e.g., BMI/treatment status
:param str classType: Type of data in ``classes``, either 'Plot Sample Type', 'categorical' or 'continuous', must be specified if classes is not ``None``. If ``classType`` is 'Plot Sample Type', ``classes`` expects 'Study Sample', 'Study Reference', 'Long-Term Reference', 'Serial Dilution' or 'Sample'.
:param str classType: Type of data in ``classes``, either 'categorical' or 'continuous', must be specified if classes is not ``None``. If ``classType`` is 'Plot Sample Type', ``classes`` expects 'Study Sample', 'Study Reference', 'Long-Term Reference', 'Serial Dilution' or 'Sample'.
:param components: If ``None`` plots all components in model, else plots those specified in components
:type components: tuple (int, int)
:param float alpha: Significance value for plotting Hotellings ellipse
Expand All @@ -80,8 +80,18 @@ def plotScores(pcaModel, classes=None, classType=None, components=None, alpha =
if not isinstance(pcaModel, ChemometricsPCA):
raise TypeError('PCAmodel must be an instance of ChemometricsPCA')

if classes is not None and classType is None:
raise ValueError('If classes is specified, classType must be')
if classes is not None and colourType is None:
raise ValueError('If classes is specified, colourType must be')

if colourType:
if colourType not in {'categorical', 'continuous', 'continuousCentered'}:
raise ValueError('colourType must be == ' + str({'categorical', 'continuous'}))

# If colourDict check colour defined for every unique entry in class
if classes is not None and colourDict is not None:
uniq = classes.unique()
if not all(k in colourDict.keys() for k in uniq):
raise ValueError('If classes and colourDict are specified every unique entry in class must be a key in colourDict')

from matplotlib.patches import Ellipse

Expand All @@ -108,9 +118,9 @@ def plotScores(pcaModel, classes=None, classType=None, components=None, alpha =

if classes is None:
classes = pandas.Series('Sample' for i in range(ns))
classType = 'categorical'
colourType = 'categorical'

if classType == 'categorical':
if colourType == 'categorical':
classes = classes.astype(str)

uniq = classes.unique()
Expand Down Expand Up @@ -155,73 +165,43 @@ def plotScores(pcaModel, classes=None, classType=None, components=None, alpha =
ax.set_xlim([(xmin + (0.2 * xmin)), xmax + (0.2 * xmax)])
ax.set_ylim([(ymin + (0.2 * ymin)), ymax + (0.2 * ymax)])

# If colouring by Sample Type - use standard reporting colours
if classType == 'Plot Sample Type':

## Try loading toolbox wide color scheme
# value just in case
sTypeColourDict = {SampleType.StudySample: 'b', SampleType.StudyPool: 'g', SampleType.ExternalReference: 'r',
SampleType.MethodReference: 'm', SampleType.ProceduralBlank: 'c', 'Other': 'grey'}
# load from the SOP as we do not have access to object
try:
from .._toolboxPath import toolboxPath
import json
import os
import copy

with open(os.path.join(toolboxPath(), 'StudyDesigns', 'SOP', 'Generic.json')) as data_file:
attributes = json.load(data_file)
# convert key names to SampleType enum
if 'sampleTypeColours' in attributes.keys():
sTypeColourDict = copy.deepcopy(attributes['sampleTypeColours'])
for stype in SampleType:
if stype.name in sTypeColourDict.keys():
sTypeColourDict[stype] = sTypeColourDict.pop(stype.name)
except:
pass

sns.set_color_codes(palette='deep')
if any(classes == 'Study Sample'):
ax.scatter(values[classes.values == 'Study Sample', components[i]], values[classes.values == 'Study Sample', components[j]], c=sTypeColourDict[SampleType.StudySample], label='Study Sample')
if any(classes == 'Study Reference'):
ax.scatter(values[classes.values == 'Study Reference', components[i]], values[classes.values == 'Study Reference', components[j]], c=sTypeColourDict[SampleType.StudyPool], label='Study Reference')
if any(classes == 'Long-Term Reference'):
ax.scatter(values[classes.values == 'Long-Term Reference', components[i]], values[classes.values == 'Long-Term Reference', components[j]], c=sTypeColourDict[SampleType.ExternalReference], label='Long-Term Reference')
if any(classes == 'Linearity Reference'):
ax.scatter(values[classes.values == 'Serial Dilution', components[i]], values[classes.values == 'Serial Dilution', components[j]], c=sTypeColourDict[SampleType.MethodReference], label='Serial Dilution')
if any(classes == 'Sample'):
ax.scatter(values[classes.values == 'Sample', components[i]], values[classes.values == 'Sample', components[j]], c=sTypeColourDict['Other'], label='Sample')
ax.legend(loc='upper left', bbox_to_anchor=(1, 1))
if colourType == 'categorical':

elif classType == 'categorical':

colors_sns = {}
# First plot any nans
if any(u in {'nan', 'NaN', 'NaT', '', 'NA'} for u in uniq):
nans = [i for i, x in enumerate(classes) if x in {'nan', 'NaN', 'NaT', '', 'NA'}]
ax.scatter(values[nans, components[i]], values[nans, components[j]], c='#D3D3D3', label='NA')
nans = [u in {'nan', 'NaN', 'NaT', '', 'NA'} for u in uniq]
nans = [i for i, x in enumerate(uniq) if x not in {'nan', 'NaN', 'NaT', '', 'NA'}]
uniqnonan = uniq[nans]
colors_sns['NA'] = '#D3D3D3'
# Plot according to user defined colours if available
if colourDict is not None:
for u in uniq:
ax.scatter(values[classes.values == u, components[i]],
values[classes.values == u, components[j]],
c=colourDict[u], label=u)

else:
uniqnonan = uniq

# Plot remaining categories
classIX=0
colors = iter(plt.cm.rainbow(numpy.linspace(0,1,len(uniqnonan))))
for u in uniqnonan:
c = rgb2hex(next(colors))
if classIX<20:
ax.scatter(values[classes.values == u, components[i]], values[classes.values == u, components[j]], c=c, label=u)#olors[classIX], label=u)
elif classIX==len(uniqnonan)-1:
ax.scatter(values[classes.values == u, components[i]], values[classes.values == u, components[j]], c='0', alpha=0, label='...')
ax.scatter(values[classes.values == u, components[i]], values[classes.values == u, components[j]], c=c, label=u)#colors[classIX], label=u)
colors_sns = {}

# First plot any nans
if any(u in {'nan', 'NaN', 'NaT', '', 'NA'} for u in uniq):
nans = [i for i, x in enumerate(classes) if x in {'nan', 'NaN', 'NaT', '', 'NA'}]
ax.scatter(values[nans, components[i]], values[nans, components[j]], c='#D3D3D3', label='NA')
nans = [i for i, x in enumerate(uniq) if x not in {'nan', 'NaN', 'NaT', '', 'NA'}]
uniqnonan = uniq[nans]
colors_sns['NA'] = '#D3D3D3'

else:
ax.scatter(values[classes.values == u, components[i]], values[classes.values == u, components[j]], c=c, label='_nolegend_')#colors[classIX], label='_nolegend_')
classIX = classIX+1
colors_sns[u] = c
uniqnonan = uniq

# Then plot remaining classes using rainbow colourmap
classIX=0
colors = iter(plt.cm.rainbow(numpy.linspace(0,1,len(uniqnonan))))
for u in uniqnonan:
c = rgb2hex(next(colors))
if classIX<20:
ax.scatter(values[classes.values == u, components[i]], values[classes.values == u, components[j]], c=c, label=u)#olors[classIX], label=u)
elif classIX==len(uniqnonan)-1:
ax.scatter(values[classes.values == u, components[i]], values[classes.values == u, components[j]], c='0', alpha=0, label='...')
ax.scatter(values[classes.values == u, components[i]], values[classes.values == u, components[j]], c=c, label=u)#colors[classIX], label=u)
else:
ax.scatter(values[classes.values == u, components[i]], values[classes.values == u, components[j]], c=c, label='_nolegend_')#colors[classIX], label='_nolegend_')
classIX = classIX+1
colors_sns[u] = c

if plotAssociation is not None:

Expand All @@ -244,7 +224,7 @@ def plotScores(pcaModel, classes=None, classType=None, components=None, alpha =

ax.legend(loc='upper left', bbox_to_anchor=(1, 1))

elif classType == 'continuous':
elif colourType == 'continuous':

plotnans = classes.isnull().values
if sum(plotnans != 0):
Expand Down
Loading

0 comments on commit 7c774c9

Please sign in to comment.