Merge pull request #74 from poldrack/develop

Develop

.circleci/config.yml

@@ -62,6 +62,6 @@ jobs:
       - run:
           name: create results tarball
           command: |
-             cd /tmp/data; tar zcvf results.tgz ./output/npboot_output.RData ./cached ./figures ./logs ./metadata ./image_diagnostics* ./output/unthresh_concat_zstat/hypo*_voxelmap.nii.gz ./output/overlap_binarized_thresh/hypo*.nii.gz ./output/consensus_analysis/hypo*_1-fdr.nii.gz ./output/ALE/hypo*_fdr_oneminusp.nii.gz ./output/cluster_maps ./output/correlation_unthresh ./PredictionMarkets
+             cp /root/project/ImageAnalyses/README_outputs.txt /tmp/data; cd /tmp/data; tar zcvf results.tgz ./output/npboot_output.RData ./cached ./figures ./logs ./metadata ./image_diagnostics* ./output/unthresh_concat_zstat/hypo*_voxelmap.nii.gz ./output/overlap_binarized_thresh/hypo*.nii.gz ./output/consensus_analysis/hypo*_1-fdr.nii.gz ./output/ALE/hypo*_fdr_oneminusp.nii.gz ./output/cluster_maps ./output/correlation_unthresh ./PredictionMarkets ./README_outputs.txt
       - store_artifacts:
              path: /tmp/data/results.tgz

ImageAnalyses/README_outputs.txt

@@ -0,0 +1,154 @@
+Outputs from automated analysis code for NARPS project
+See https://github.com/poldrack/narps for more details
+
+Legend for files contained in this directory:
+
+
+./image_diagnostics_zstat: image diagnostics for resampled statistical images (one file per team)
+
+./image_diagnostics_orig: image diagnostics for original (submitted) statistical images (one file per team)
+
+./output: outputs from analyses
+
+./output/npboot_output.RData: stored samples from nonparametric bootstrap analysis
+
+./output/overlap_binarized_thresh: images showing proportion of teams with activation at each voxel (separate image for each hypothesis)
+
+./output/ALE: outputs from coordinate-based meta-analysis, containing 1 - FDR-corrected p-value (separate image for each hypothesis)
+
+./output/cluster_maps: average unthresholded statistical maps for clusters of teams (separate maps for each hypothesis/cluster)
+
+./output/consensus_analysis: outputs from image-based meta-analysis, containing 1 - FDR-corrected p-value (separate image for each hypothesis)
+
+./output/correlation_unthresh: correlation data for unthresholded maps
+./output/correlation_unthresh/spearman_unthresh_hyp[1-9].csv: correlation matrix for unthresholded maps between teams for each hypothesis
+./output/correlation_unthresh/unthresh_cluster_membership_spearman.json: stored cluster membership
+./output/correlation_unthresh/mean_unthresh_correlation_by_cluster.csv: mean correlation between teams within each cluster
+
+./output/unthresh_concat_zstat/hypo[1-9]_voxelmap.nii.gz: maps showing proportion of teams with data at each voxel
+
+./cached/narps_prepare_maps.pkl: cached data structure for NARPS analysis
+
+./logs/AnalyzeMaps-plot_distance_from_mean.txt:  results from analysis of distance from mean pattern 
+
+./logs/AnalyzeMaps-mk_correlation_maps_unthresh.txt: results from analysis of number of nonzero voxels (unthresh zstat) and median correlation between teams 
+
+./logs/narps.txt: main log file for analysis
+
+./logs/ThresholdSimulation.log: results from simulation using consistent thresholding 
+
+./logs/AnalyzeMaps-analyze_clusters.txt: results from cluster-specific analyses of unthresholded maps
+
+./logs/AnalyzeMaps-get_thresh_similarity.txt: results from analysis of percent agreement and Jaccard similarity across thresholded maps
+
+./logs/image_diagnostics.log: log from image processing 
+
+./logs/ConsensusAnalysis.txt: results from image-based meta-analysis, including tau statistics 
+
+./logs/AnalyzeMaps-mk_overlap_maps.txt: results from analysis of maximum voxel overlap 
+
+./logs/zstat_diagnostics.log: log from processing of zstat maps 
+
+./figures: analysis outputs and figures (note: for many of these figures there are both PDF and PNG versions present)
+
+./figures/hyp[1-9]_combined_clusters.png: combined heatmap and cluster slices (separately for each hypothesis)
+
+./figures/hyp[1-9]_pctagree_map_thresh.pdf: heatmaps for percent agreement
+
+./figures/ModelingSummaryTable.tsv: summary of results from mixed effects logistic regression and nonparametric bootstrap analyses
+./figures/OddsRatios.tsv
+
+./figures/hyp[1-9]_ALE_z.png: Z statistic maps for coordinate-based meta-analysis
+./figures/hyp[1-9]_ALE_p.png: uncorrected p-value maps for coordinate-based meta-analysis
+./figures/hyp[1-9]_ALE_fdr_oneminusp.png: FDR corrected p-value maps for coordinate-based meta-analysis
+./figures/hyp[1-9]_ALE_fdr_thresh_0.05.png: thresholded FDR corrected p-value maps for coordinate-based meta-analysis
+./figures/hyp[1-9]_ALE_ale.png: Activation likelihood estimation maps for coordinate-based meta-analysis 
+./figures/ALE_map.png
+
+./figures/hyp[1-9]_spearman_map_unthresh.pdf: heatmap for spearman correlation of unthresholded maps between teams
+
+./figures/range_map.png: maps of range in Z stats for all hypotheses
+
+./figures/cluster_correlation.pdf: correlation map between all hypotheses/clusters
+
+./figures/overlap_map.png: maps of overlap in thresholded maps for all hypotheses
+
+./figures/hyp[1-9]_cluster_means.png: thresholded mean activation map for each cluster, separately for each hypothesis
+
+./figures/tau_histograms.pdf: histograms of tau statistic across voxels for all hypotheses
+
+./figures/tau_maps.pdf: tau maps for all hypotheses
+
+./figures/ConfidenceDataWide.csv: confidence data for each team (used in generating final table)
+
+./figures/median_corr_sorted.pdf: sorted median correlation for each team
+
+./figures/std_map.png: maps of standard deviation across z statistics for all hypotheses
+
+./figures/Table1.tsv: Summary of team reports (decision, confidence, similarity)
+
+./figures/TomEtAl_correlation.pdf: correlation of each hypothesis/cluster with maps from original Tom et al. study
+
+./figures/SuppFigure1.png: original version of figure showing all teams with modeling choices
+
+./figures/ThresholdSimulation: results from simulations using consistent thresholding
+./figures/ThresholdSimulation/simulation_results.csv: summary of overall results
+./figures/ThresholdSimulation/ventralstriatum_mask.nii.gz: mask for ventral striatum
+./figures/ThresholdSimulation/vmpfc_mask.nii.gz: mask for vmpfc
+./figures/ThresholdSimulation/amygdala_mask.nii.gz: mask for amygdala
+./figures/ThresholdSimulation/decision_vs_activation.png: plot of relative decision likelihood for original vs. re-thresholded data
+
+./figures/MethodsTableMetadataMerged.csv: analysis metadata, using in creating final table
+
+./figures/DecisionAnalysis.html: results from logistic regression analyses and bootstrap
+
+./figures/NARPS_mean_correlation.pdf: correlation for each hypothesis/cluster with overall activation map from original NARPS analysis
+
+./figures/DecisionDataWide.csv: decision data (used to generate final table)
+
+./figures/consensus_map.pdf: maps from image-based meta-analysis
+
+./figures/correlation_unthresh
+./figures/correlation_unthresh/spearman_unthresh_hyp[1-9]_cluster[1-3].png: histograms of correlations for each hypothesis/cluster
+./figures/correlation_unthresh/spearman_unthresh_hyp[1-9]_mean.png: histograms of correlations for each hypothesis overall
+
+./PredictionMarkets: results from prediction market analyses
+./PredictionMarkets/Figures/SuppTable_HoldingStats.tsv: Analyses of holdings
+./PredictionMarkets/Figures/Active_vs_All_traders.pdf: Figure showing analysis of all traders vs. active traders
+./PredictionMarkets/Figures/SuppTable_MarketDetails.tsv: Details regarding markets
+./PredictionMarkets/Figures/SuppTable_MarketResults.tsv: Comparison of team vs. non-team market results
+./PredictionMarkets/Figures/SuppTable_PanelRegressions.tsv: results from regression models on market results
+./PredictionMarkets/Figures/PMbeliefs_Figure3.pdf: figure showing final market price vs. fundamental value
+./PredictionMarkets/Figures/PM_Analyses.html: Results from Prediction market analyses
+./PredictionMarkets/Figures/PM_Figures.html: Output from figure creation notebook
+./PredictionMarkets/Figures/Timeseries.pdf: Price timeseries for each market
+
+./PredictionMarkets/Combined: Data files used in Prediction Market analyses
+./PredictionMarkets/Combined/position.RData
+./PredictionMarkets/Combined/asset.RData
+./PredictionMarkets/Combined/portfolio.RData
+./PredictionMarkets/Combined/transaction.RData
+./PredictionMarkets/Combined/teamresults.RData
+
+./PredictionMarkets/Processed: Intermediate data files used in Prediction Market analyses
+./PredictionMarkets/Processed/Fundamentals.RData
+./PredictionMarkets/Processed/Prices.RData
+./PredictionMarkets/Processed/BalancedPanel.RData
+./PredictionMarkets/Processed/Transactions.RData
+./PredictionMarkets/Processed/TeamResults.RData
+./PredictionMarkets/Processed/Holdings.RData
+
+./metadata: various metadata outputs
+./metadata/cluster_corr_TomEtAl.csv: correlation between each hypothesis/cluster and Tom et al. original maps
+./metadata/cluster_metadata_df.csv: cluster membership by team/hypothesis
+./metadata/thresh_voxel_statistics.csv: statistics on thresholded maps for each hypothesis
+./metadata/smoothness_est.csv: smoothness estimates per hypothesis/team
+./metadata/all_metadata.csv: full metadata output
+./metadata/rectified_images_list.txt: list of teams/hypotheses that required direction-flipping
+./metadata/cluster_corr_NARPS_mean.csv: correlation of each hypothesis/cluster with NARPS overall task activation maps
+./metadata/pctagree_hyp[1-9].csv: percent agreement data across teams
+./metadata/image_metadata_df.csv: number of NA and nonzero voxels for each team
+./metadata/narps_metadata_all_teams.csv: original metadata for all teams
+./metadata/median_pattern_corr.csv: median correlation for each team with mean map across teams
+./metadata/thresh_voxel_data.csv: metadata for each team's thresholded maps
+./metadata/R_package_info.txt: listing of all R package/versions used in analysis