Summary: weekly mtg 20170412 (Byron, Steve, Matt, Aaron, Emily, me) - mobeets/nullSpaceControl GitHub Wiki

• Frequency -> # of occurrences
• Spikes/timestep -> Hz
• remove "Activity (spikes/timestep)"
• verify scale bar
• strategy: label the hell out of Fig. 1D so readers are prepared

Fig. 1D

• label points A and B
• need to link to v_1, like "v = v_1"
• remove tick mark
• maybe remove C? or color it differently?
  • remove it; can just ref. the idea that to produce a different velocity you would need to change the activity along the output-potent axis
• maybe have D be two panels? one for v_1, one for v_2
• make sure xlabel is "Activity (spikes/second)", and title is "Output null dim."

Fig. 1C

• spikes/timestep -> spikes/second
• maybe get rid of "Cursor velocity", but then might want to link Cursor vel. to Output-potent somewhere else? it's maybe just not connected to panel C.

Fig. 2B

• xlabel: "Activity (spikes/second)"
• xticks: μ-5, μ, μ+5
  • or can just have three unlabeled ticks/dots, and mention what they are in the axis
• Or: xticks: -5, 0, 5, with label being spikes/sec, relative to baseline
  • no "Activity", just numbers as ticks, and label is the units "spikes/sec rel. to baseline"
• scale bar only in 2C
• rename title "Output-null dim. 1 activity"

Fig. 2C

• just mention in caption that the error refers to across session
• bigger text for error message

Fig. 2

• make hypothesis titles slightly above each column, not just in A
• maybe diff. colors for cloud vs. hab?

Fig. 1D: should look exactly like Fig. 2C

• center histograms over the points
• have Fig. 1C with a point D to go with C
• loopy arrow pointing to each of the two axes

Say in Fig. 1C caption that where 0 is is arbitrary, and we define by convention that 0 is the mean activity across the session or timesteps or whatever. also need to define what "baseline" means in the figure caption and again in methods.