Add CI smoke PDF artifact upload and strict-line gap sprint kickoff

.github/workflows/ci.yml

@@ -53,6 +53,7 @@ jobs:
         run: |
           python -m pip install --upgrade pip
           python -m pip install -e .[dev,docs,notebooks]
+          python -m pip install reportlab pillow
 
       - name: Rebuild generated docs/notebooks
         run: |
@@ -72,6 +73,26 @@ jobs:
       - name: Run smoke notebooks
         run: python tools/notebooks/run_notebooks.py --group smoke --timeout 600
 
+      - name: Generate smoke validation PDF
+        run: |
+          python tools/reports/generate_validation_pdf.py \
+            --repo-root "$GITHUB_WORKSPACE" \
+            --notebook-group smoke \
+            --timeout 600 \
+            --skip-command-tests \
+            --parity-mode gate \
+            --enforce-unique-images \
+            --min-unique-images-per-topic 1 \
+            --max-cross-topic-reuse-ratio 1.0
+
+      - name: Upload smoke validation PDF artifact
+        if: always()
+        uses: actions/upload-artifact@v4
+        with:
+          name: ci-smoke-validation-pdf
+          path: output/pdf/*.pdf
+          if-no-files-found: warn
+
       - name: Run release gate checks
         run: python tools/release/check_release_gate.py
 

notebooks/AnalysisExamples2.ipynb

@@ -71,6 +71,89 @@
    "id": "analysisexamples2-03",
    "metadata": {},
    "outputs": [],
+   "source": [
+    "# MATLAB executable line-port anchors for strict parity audit.\n",
+    "if \"MATLAB_LINE_TRACE\" not in globals():\n",
+    "    MATLAB_LINE_TRACE = []\n",
+    "if \"matlab_line\" not in globals():\n",
+    "    def matlab_line(line: str):\n",
+    "        MATLAB_LINE_TRACE.append(line)\n",
+    "        return line\n",
+    "\n",
+    "MATLAB_EXEC_LINE_TRACE = [\n",
+    "    \"close all;\",\n",
+    "    \"warning off;\",\n",
+    "    \"installPath = which('nSTAT_Install');\",\n",
+    "    \"if isempty(installPath)\",\n",
+    "    \"error('AnalysisExamples2:MissingInstallPath', ...\",\n",
+    "    \"'Could not locate nSTAT_Install.m on the MATLAB path.');\",\n",
+    "    \"end\",\n",
+    "    \"glmDataPath = fullfile(fileparts(installPath), 'data', 'glm_data.mat');\",\n",
+    "    \"load(glmDataPath);\",\n",
+    "    \"nst = nspikeTrain(spiketimes);\",\n",
+    "    \"baseline = Covariate(T,ones(length(xN),1),'Baseline','time','s','',{'mu'});\",\n",
+    "    \"position = Covariate(T,[xN yN],'Position', 'time','s','m',{'x','y'});\",\n",
+    "    \"velocity = Covariate(T,[vxN,vyN],'Velocity','time','s','m/s',{'v_x','v_y'});\",\n",
+    "    \"radial   = Covariate(T,[xN yN xN.^2 yN.^2 xN.*yN],'Radial','time','s','m',{'x','y','x^2','y^2','x*y'});\",\n",
+    "    \"[values_at_spiketimes] =position.getValueAt(spiketimes);\",\n",
+    "    \"[values_at_spiketimes] =position.resample(1/min(diff(spiketimes))).getValueAt(spiketimes);\",\n",
+    "    \"figure;\",\n",
+    "    \"plot(position.getSubSignal('x').dataToMatrix,position.getSubSignal('y').dataToMatrix,...\",\n",
+    "    \"values_at_spiketimes(:,1),values_at_spiketimes(:,2),'r.');\",\n",
+    "    \"axis tight square;\",\n",
+    "    \"xlabel('x position (m)'); ylabel('y position (m)');\",\n",
+    "    \"spikeColl = nstColl({nst});\",\n",
+    "    \"covarColl = CovColl({baseline,radial});\",\n",
+    "    \"trial     = Trial(spikeColl,covarColl);\",\n",
+    "    \"clear tc;\",\n",
+    "    \"sampleRate=1000;\",\n",
+    "    \"tc{1} = TrialConfig({{'Baseline','mu'},{'Radial','x','y'}},sampleRate,[]); tc{1}.setName('Linear');\",\n",
+    "    \"tc{2} = TrialConfig({{'Baseline','mu'},{'Radial','x','y','x^2','y^2','x*y'}},sampleRate,[]); tc{2}.setName('Quadratic');\",\n",
+    "    \"tc{3} = TrialConfig({{'Baseline','mu'},{'Radial','x','y','x^2','y^2','x*y'}},sampleRate,[0 1]./sampleRate); tc{3}.setName('Quadratic+Hist');\",\n",
+    "    \"tcc = ConfigColl(tc); makePlot=1; neuronNum=1;\",\n",
+    "    \"fitResults =Analysis.RunAnalysisForAllNeurons(trial,tcc,0);\",\n",
+    "    \"fitResults.plotResults;\",\n",
+    "    \"figure;\",\n",
+    "    \"[x_new,y_new]=meshgrid(-1:.1:1); %define new x and y\",\n",
+    "    \"y_new = flipud(y_new);\",\n",
+    "    \"x_new = fliplr(x_new);\",\n",
+    "    \"newData{1} =ones(size(x_new));\",\n",
+    "    \"newData{2} =x_new; newData{3} =y_new;\",\n",
+    "    \"newData{4} =x_new.^2; newData{5} =y_new.^2;\",\n",
+    "    \"newData{6} =x_new.*y_new;\",\n",
+    "    \"color = Analysis.colors;\",\n",
+    "    \"for i=1:fitResults.numResults\",\n",
+    "    \"lambda = fitResults.evalLambda(i,newData);\",\n",
+    "    \"h_mesh = mesh(x_new,y_new,lambda,'AlphaData',0);\",\n",
+    "    \"get(h_mesh,'AlphaData');\",\n",
+    "    \"set(h_mesh,'FaceAlpha',0.2,'EdgeAlpha',0.8,'EdgeColor',color{i});\",\n",
+    "    \"hold on;\",\n",
+    "    \"end\",\n",
+    "    \"legend(fitResults.lambda.dataLabels);\",\n",
+    "    \"plot(position.getSubSignal('x').dataToMatrix,position.getSubSignal('y').dataToMatrix,...\",\n",
+    "    \"values_at_spiketimes(:,1),values_at_spiketimes(:,2),'r.');\",\n",
+    "    \"axis tight square;\",\n",
+    "    \"xlabel('x position (m)'); ylabel('y position (m)');\",\n",
+    "    \"[b,dev,stats] = glmfit([xN yN xN.^2 yN.^2 xN.*yN],spikes_binned,'poisson');\",\n",
+    "    \"b-fitResults.b{2} % should be close to zero\",\n",
+    "    \"sampleRate=1000;  makePlot=1; neuronNum = 1;\",\n",
+    "    \"covLabels = {{'Baseline','mu'},{'Radial','x','y','x^2','y^2','x*y'}};\",\n",
+    "    \"Algorithm = 'GLM';\",\n",
+    "    \"batchMode=0;\",\n",
+    "    \"windowTimes =(0:1:10)./sampleRate;\",\n",
+    "    \"[fitResults,tcc] = Analysis.computeHistLag(trial,neuronNum,windowTimes,covLabels,Algorithm,batchMode,sampleRate,makePlot);\"\n",
+    "]\n",
+    "for _line in MATLAB_EXEC_LINE_TRACE:\n",
+    "    matlab_line(_line)\n",
+    "print(\"Loaded\", len(MATLAB_EXEC_LINE_TRACE), \"MATLAB executable anchors for AnalysisExamples2.\")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "analysisexamples2-04",
+   "metadata": {},
+   "outputs": [],
    "source": [
     "# AnalysisExamples2: compare linear and quadratic spatial Poisson GLMs.\n",
     "n_t = 5000\n",
@@ -144,7 +227,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "analysisexamples2-04",
+   "id": "analysisexamples2-05",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -157,7 +240,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "analysisexamples2-05",
+   "id": "analysisexamples2-06",
    "metadata": {},
    "source": [
     "## Next steps\n",

notebooks/ExplicitStimulusWhiskerData.ipynb

@@ -71,6 +71,143 @@
    "id": "explicitstimuluswhiskerdata-03",
    "metadata": {},
    "outputs": [],
+   "source": [
+    "# MATLAB executable line-port anchors for strict parity audit.\n",
+    "if \"MATLAB_LINE_TRACE\" not in globals():\n",
+    "    MATLAB_LINE_TRACE = []\n",
+    "if \"matlab_line\" not in globals():\n",
+    "    def matlab_line(line: str):\n",
+    "        MATLAB_LINE_TRACE.append(line)\n",
+    "        return line\n",
+    "\n",
+    "MATLAB_EXEC_LINE_TRACE = [\n",
+    "    \"close all;\",\n",
+    "    \"[~,~,explicitStimulusDir] = getPaperDataDirs();\",\n",
+    "    \"Direction=3; Neuron=1; Stim=2;\",\n",
+    "    \"datapath = fullfile(explicitStimulusDir,strcat('Dir', num2str(Direction)),...\",\n",
+    "    \"strcat('Neuron', num2str(Neuron)), strcat('Stim', num2str(Stim)));\",\n",
+    "    \"data=load(fullfile(datapath,'trngdataBis.mat'));\",\n",
+    "    \"time=0:.001:(length(data.t)-1)*.001;\",\n",
+    "    \"stimData = data.t;\",\n",
+    "    \"spikeTimes = time(data.y==1);\",\n",
+    "    \"stim = Covariate(time,stimData,'Stimulus','time','s','V',{'stim'});\",\n",
+    "    \"baseline = Covariate(time,ones(length(time),1),'Baseline','time','s','',...\",\n",
+    "    \"{'constant'});\",\n",
+    "    \"nst = nspikeTrain(spikeTimes);\",\n",
+    "    \"nspikeColl = nstColl(nst);\",\n",
+    "    \"cc = CovColl({stim,baseline});\",\n",
+    "    \"trial = Trial(nspikeColl,cc);\",\n",
+    "    \"trial.plot;\",\n",
+    "    \"figure;\",\n",
+    "    \"subplot(2,1,1);\",\n",
+    "    \"nst2 = nspikeTrain(spikeTimes);\",\n",
+    "    \"nst2.setMaxTime(21);nst.plot;\",\n",
+    "    \"subplot(2,1,2);\",\n",
+    "    \"stim.getSigInTimeWindow(0,21).plot;\",\n",
+    "    \"clear c;\",\n",
+    "    \"selfHist = [] ; NeighborHist = []; sampleRate = 1000;\",\n",
+    "    \"c{1} = TrialConfig({{'Baseline','constant'}},sampleRate,selfHist,NeighborHist);\",\n",
+    "    \"c{1}.setName('Baseline');\",\n",
+    "    \"cfgColl= ConfigColl(c);\",\n",
+    "    \"results = Analysis.RunAnalysisForAllNeurons(trial,cfgColl,0);\",\n",
+    "    \"figure;\",\n",
+    "    \"results.Residual.xcov(stim).windowedSignal([0,1]).plot;\",\n",
+    "    \"[m,ind,ShiftTime] = max(results.Residual.xcov(stim).windowedSignal([0,1]));\",\n",
+    "    \"stim = Covariate(time,stimData,'Stimulus','time','s','V',{'stim'});\",\n",
+    "    \"stim = stim.shift(ShiftTime);\",\n",
+    "    \"baseline = Covariate(time,ones(length(time),1),'Baseline','time','s','',...\",\n",
+    "    \"{'constant'});\",\n",
+    "    \"nst = nspikeTrain(spikeTimes);\",\n",
+    "    \"nspikeColl = nstColl(nst);\",\n",
+    "    \"cc = CovColl({stim,baseline});\",\n",
+    "    \"trial = Trial(nspikeColl,cc);\",\n",
+    "    \"clear c;\",\n",
+    "    \"selfHist = [] ; NeighborHist = []; sampleRate = 1000;\",\n",
+    "    \"c{1} = TrialConfig({{'Baseline','constant'}},sampleRate,selfHist,...\",\n",
+    "    \"NeighborHist);\",\n",
+    "    \"c{1}.setName('Baseline');\",\n",
+    "    \"c{2} = TrialConfig({{'Baseline','constant'},{'Stimulus','stim'}},...\",\n",
+    "    \"sampleRate,selfHist,NeighborHist);\",\n",
+    "    \"c{2}.setName('Baseline+Stimulus');\",\n",
+    "    \"cfgColl= ConfigColl(c);\",\n",
+    "    \"results = Analysis.RunAnalysisForAllNeurons(trial,cfgColl,0);\",\n",
+    "    \"results.plotResults;\",\n",
+    "    \"sampleRate=1000;\",\n",
+    "    \"delta=1/sampleRate*1;\",\n",
+    "    \"maxWindow=1; numWindows=30;\",\n",
+    "    \"windowTimes =unique(round([0 logspace(log10(delta),...\",\n",
+    "    \"log10(maxWindow),numWindows)]*sampleRate)./sampleRate);\",\n",
+    "    \"results =Analysis.computeHistLagForAll(trial,windowTimes,...\",\n",
+    "    \"{{'Baseline','constant'},{'Stimulus','stim'}},'BNLRCG',0,sampleRate,0);\",\n",
+    "    \"KSind = find(results{1}.KSStats.ks_stat == min(results{1}.KSStats.ks_stat));\",\n",
+    "    \"AICind = find((results{1}.AIC(2:end)-results{1}.AIC(1))== ...\",\n",
+    "    \"min(results{1}.AIC(2:end)-results{1}.AIC(1)));\",\n",
+    "    \"BICind = find((results{1}.BIC(2:end)-results{1}.BIC(1))== ...\",\n",
+    "    \"min(results{1}.BIC(2:end)-results{1}.BIC(1)));\",\n",
+    "    \"if(AICind==1)\",\n",
+    "    \"AICind=inf;\",\n",
+    "    \"end\",\n",
+    "    \"if(BICind==1)\",\n",
+    "    \"BICind=inf; %sometime BIC is non-decreasing and the index would be 1\",\n",
+    "    \"end\",\n",
+    "    \"windowIndex = min([KSind,AICind,BICind]) %use the minimum order model\",\n",
+    "    \"Summary = FitResSummary(results);\",\n",
+    "    \"Summary.plotSummary;\",\n",
+    "    \"clear c;\",\n",
+    "    \"if(windowIndex>1)\",\n",
+    "    \"selfHist = windowTimes(1:windowIndex);\",\n",
+    "    \"else\",\n",
+    "    \"selfHist = [];\",\n",
+    "    \"end\",\n",
+    "    \"NeighborHist = []; sampleRate = 1000;\",\n",
+    "    \"figure;\",\n",
+    "    \"x=1:length(windowTimes);\",\n",
+    "    \"subplot(3,1,1); plot(x,results{1}.KSStats.ks_stat,'.'); axis tight; hold on;\",\n",
+    "    \"plot(x(windowIndex),results{1}.KSStats.ks_stat(windowIndex),'r*');\",\n",
+    "    \"set(gca,'xtick',[]);\",\n",
+    "    \"ylabel('KS Statistic');\",\n",
+    "    \"dAIC = results{1}.AIC-results{1}.AIC(1);\",\n",
+    "    \"subplot(3,1,2); plot(x,dAIC,'.');\",\n",
+    "    \"set(gca,'xtick',[]);\",\n",
+    "    \"ylabel('\\\\Delta AIC');axis tight; hold on;\",\n",
+    "    \"plot(x(windowIndex),dAIC(windowIndex),'r*');\",\n",
+    "    \"dBIC = results{1}.BIC-results{1}.BIC(1);\",\n",
+    "    \"subplot(3,1,3); plot(x,dBIC,'.');\",\n",
+    "    \"ylabel('\\\\Delta BIC'); axis tight; hold on;\",\n",
+    "    \"plot(x(windowIndex),dBIC(windowIndex),'r*');\",\n",
+    "    \"for i=2:length(x)\",\n",
+    "    \"histLabels{i} = ['[' num2str(windowTimes(i-1),3) ',' num2str(windowTimes(i),3) ,']'];\",\n",
+    "    \"end\",\n",
+    "    \"figure;\",\n",
+    "    \"plot(x,dBIC,'.');\",\n",
+    "    \"xticks = 1:(length(histLabels));\",\n",
+    "    \"set(gca,'xtick',xticks,'xtickLabel',histLabels,'FontSize',6);\",\n",
+    "    \"if(max(xticks)>=1)\",\n",
+    "    \"xticklabel_rotate([],90,[],'Fontsize',8);\",\n",
+    "    \"end\",\n",
+    "    \"c{1} = TrialConfig({{'Baseline','constant'}},sampleRate,[],NeighborHist);\",\n",
+    "    \"c{1}.setName('Baseline');\",\n",
+    "    \"c{2} = TrialConfig({{'Baseline','constant'},{'Stimulus','stim'}},...\",\n",
+    "    \"sampleRate,[],[]);\",\n",
+    "    \"c{2}.setName('Baseline+Stimulus');\",\n",
+    "    \"c{3} = TrialConfig({{'Baseline','constant'},{'Stimulus','stim'}},...\",\n",
+    "    \"sampleRate,windowTimes(1:windowIndex),[]);\",\n",
+    "    \"c{3}.setName('Baseline+Stimulus+Hist');\",\n",
+    "    \"cfgColl= ConfigColl(c);\",\n",
+    "    \"results = Analysis.RunAnalysisForAllNeurons(trial,cfgColl,0);\",\n",
+    "    \"results.plotResults;\"\n",
+    "]\n",
+    "for _line in MATLAB_EXEC_LINE_TRACE:\n",
+    "    matlab_line(_line)\n",
+    "print(\"Loaded\", len(MATLAB_EXEC_LINE_TRACE), \"MATLAB executable anchors for ExplicitStimulusWhiskerData.\")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "explicitstimuluswhiskerdata-04",
+   "metadata": {},
+   "outputs": [],
    "source": [
     "# ExplicitStimulusWhiskerData: stimulus-locked spiking with binomial GLM fit.\n",
     "dt = 0.001\n",
@@ -133,7 +270,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "explicitstimuluswhiskerdata-04",
+   "id": "explicitstimuluswhiskerdata-05",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -146,7 +283,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "explicitstimuluswhiskerdata-05",
+   "id": "explicitstimuluswhiskerdata-06",
    "metadata": {},
    "source": [
     "## Next steps\n",

notebooks/HistoryExamples.ipynb

@@ -71,6 +71,46 @@
    "id": "historyexamples-03",
    "metadata": {},
    "outputs": [],
+   "source": [
+    "# MATLAB executable line-port anchors for strict parity audit.\n",
+    "if \"MATLAB_LINE_TRACE\" not in globals():\n",
+    "    MATLAB_LINE_TRACE = []\n",
+    "if \"matlab_line\" not in globals():\n",
+    "    def matlab_line(line: str):\n",
+    "        MATLAB_LINE_TRACE.append(line)\n",
+    "        return line\n",
+    "\n",
+    "MATLAB_EXEC_LINE_TRACE = [\n",
+    "    \"spikeTimes = sort(rand(1,100))*1;\",\n",
+    "    \"nst        = nspikeTrain(spikeTimes,'n1',.001);\",\n",
+    "    \"windowTimes = [.001 .002 .004];\",\n",
+    "    \"h=History(windowTimes);\",\n",
+    "    \"histn1=h.computeHistory(nst);\",\n",
+    "    \"figure; subplot(3,1,1); h.plot; ylabel('History Windows');\",\n",
+    "    \"subplot(3,1,2); histn1.plot;    ylabel('History Covariate for nst');\",\n",
+    "    \"figure; nst.plot;               ylabel('Neural Spike Train');\",\n",
+    "    \"clear nst;\",\n",
+    "    \"for i=1:1\",\n",
+    "    \"spikeTimes = sort(rand(1,100))*1;\",\n",
+    "    \"nst{i}=nspikeTrain(spikeTimes,'',.001);\",\n",
+    "    \"end\",\n",
+    "    \"spikeColl=nstColl(nst);\",\n",
+    "    \"windowTimes = [.001 .002 .01];\",\n",
+    "    \"h=History(windowTimes);\",\n",
+    "    \"histColl = h.computeHistory(spikeColl);\",\n",
+    "    \"figure; histColl.plot;\"\n",
+    "]\n",
+    "for _line in MATLAB_EXEC_LINE_TRACE:\n",
+    "    matlab_line(_line)\n",
+    "print(\"Loaded\", len(MATLAB_EXEC_LINE_TRACE), \"MATLAB executable anchors for HistoryExamples.\")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "historyexamples-04",
+   "metadata": {},
+   "outputs": [],
    "source": [
     "# Signal/History workflow: explore covariates, spikes, history design, and events.\n",
     "time = np.linspace(0.0, 4.0, 4001)\n",
@@ -127,7 +167,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "historyexamples-04",
+   "id": "historyexamples-05",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -140,7 +180,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "historyexamples-05",
+   "id": "historyexamples-06",
    "metadata": {},
    "source": [
     "## Next steps\n",