Can anyone help me understand data log from a P4A?

Joined
Feb 10, 2019
Messages
8
Likes
0
Age
24
#1
Hello everyone,

I'm new to the forum and eager to find relationship between 2 or more columns (arguments, parameters?) But i need to understand what they stand for first.

Here is nearly all columns' name of a .CSV file from a Phantom 4 Advanced.

Tick# offsetTime IMU_ATTI(0):Longitude IMU_ATTI(0):Latitude IMU_ATTI(0):numSats IMU_ATTI(0):barometer:Raw IMU_ATTI(0):barometer:Smooth IMU_ATTI(0):accel:X IMU_ATTI(0):accel:Y IMU_ATTI(0):accel:Z IMU_ATTI(0):accel:Composite IMU_ATTI(0):gyro:X IMU_ATTI(0):gyro:Y IMU_ATTI(0):gyro:Z IMU_ATTI(0):gyro:Composite IMU_ATTI(0):mag:X IMU_ATTI(0):mag:Y IMU_ATTI(0):mag:Z IMU_ATTI(0):mag:Mod IMU_ATTI(0):velN IMU_ATTI(0):velE IMU_ATTI(0):velD IMU_ATTI(0):velComposite IMU_ATTI(0):velH IMU_ATTI(0):GPS-H IMU_ATTI(0):roll IMU_ATTI(0):pitch IMU_ATTI(0):yaw IMU_ATTI(0):yaw360 IMU_ATTI(0):totalGyro:Z IMU_ATTI(0):totalGyro:X IMU_ATTI(0):totalGyro:Y IMU_ATTI(0):magYaw IMU_ATTI(0):distanceHP IMU_ATTI(0):distanceTravelled IMU_ATTI(0):directionOfTravel[mag] IMU_ATTI(0):directionOfTravel[true] IMU_ATTI(0):temperature flightTime gpsHealth General:vpsHeight General:relativeHeight General:absoluteHeight GPS(0):Long GPS(0):Lat GPS(0):Date GPS(0):Time GPS:dateTimeStamp GPS(0):heightMSL GPS(0):hDOP GPS(0):pDOP GPS(0):sAcc GPS(0):numGPS GPS(0):numGLNAS GPS(0):numSV GPS(0):velN GPS(0):velE GPS(0):velD RC:Aileron RC:Elevator RC:Rudder RC:Throttle Controller:gpsLevel Controller:ctrl_level RC:failSafe RC:dataLost RC:appLost RC:sigStrength RC:connected HP:Longitude HP:Latitude HP:rthHeight Battery:status SMART_BATT:goHome% SMART_BATT:land% SMART_BATT:goHomeTime SMART_BATT:landTime OA:avoidObst OA:emergBrake OA:radiusLimit OA:airportLimit OA:groundForceLanding OA:horizNearBoundary OA:vertLowLimit OA:vertAirportLimit OA:roofLimit OA:hitGroundLimit OA:frontDistance Motor:Speed:RFront Motor:Speed:LFront Motor:Speed:LBack Motor:Speed:RBack Motor:EscTemp:RFront Motor:EscTemp:LFront Motor:EscTemp:LBack Motor:EscTemp:RBack Motor:pPMrecv:RFront Motor:pPMrecv:LFront Motor:pPMrecv:LBack Motor:pPMrecv:RBack Motor:pPMsend:RFront Motor:pPMsend:LFront Motor:pPMsend:LBack Motor:pPMsend:RBack Motor:V_out:RFront Motor:V_out:LFront Motor:V_out:LBack Motor:V_out:RBack Motor:Volts:RFront Motor:Volts:LFront Motor:Volts:LBack Motor:Volts:RBack Motor:Current:RFront Motor:Current:LFront Motor:Current:LBack Motor:Current:RBack Motor:Status:RFront Motor:Status:LFront Motor:Status:LBack Motor:Status:RBack Motor:thrustAngle AirComp:AirSpeedBody:X AirComp:AirSpeedBody:Y AirComp:Alti AirComp:WindSpeed AirComp:Wind:X AirComp:Wind:Y AirComp:MotorSpeed AirComp:VelLevel flyCState flycCommand flightAction nonGPSCause connectedToRC Battery:lowVoltage RC:ModeSwitch gpsUsed visionUsed IMUEX(0):err IMUEX(0):rtk_Longitude IMUEX(0):rtk_Latitude IMUEX(0):rtk_Alti MotorCtrl:pWM:RFront MotorCtrl:pWM:LFront MotorCtrl:pWM:LBack MotorCtrl:pWM:RBack BattInfo:Ad_v BattInfo:r_time BattInfo:AvgCurrent BattInfo:vol_t BattInfo:pack_ve BattInfo:Current BattInfo:RemainingCap BattInfo:Remaining% BattInfo:Temp BattInfo:right BattInfo:l_cell BattInfo:dyna_cnt BattInfo:FullCap BattInfo:eek:ut_ctl BattInfo:eek:ut_ctl_f AirCraftCondition:fsmState AirCraftCondition:nearGround AirCraftCondition:landState AirCraftCondition:UP_acc_t AirCraftCondition:UP_TF_t Attribute|Value ConvertDatV3 3.4.0

I can only understand a small part of them. Can you tell me which thread I should read? or each meaning? Thanks a lot.
 
Joined
Feb 10, 2019
Messages
8
Likes
0
Age
24
#3
Most of the names are self-explanatory, with a few exceptions. Which ones were you interested in?

There is a guide to the engineered (not DAT-defined) fields on the DatCon website:

V3 .CSV column descriptions
Thank you for your help. I've read P3 and Inspire 1 .CSV column descriptions before posting new thread. V3 .CSV column descriptions explains most remaining names well.

However, there are still some names that "descriptions" includes without explanations. Such as "Motor:V_out:RBack" .

Capture.PNG

PPM may refers to pulse-posion modulation but I'm not certain. Battery(Batt#) & BattInfo are the 2 parts I'm interested in if I have to choose.

Again, thanks a lot.
 
Joined
Mar 22, 2014
Messages
4,612
Likes
2,161
Location
Los Alamos, NM.
#5
Is there anyone with any suggestion for this problem?


Thanks in advance.
What, exactly, is the problem? If you are referring to the fields that you indicated in the screen shot, the IMUEX data is mostly not relevant to the Phantom, the PPM is pulse position modulation, and the motor data are motor voltage.
 
Joined
Feb 10, 2019
Messages
8
Likes
0
Age
24
#6
TY. That's the point.
Maybe I should focus on the meaning of data in the same scope (prefix), such as IMUEX,IMU_ATTI ,OA, and try to find relationships among them.

Actually, the problem is that I WAS WONDERING ABOUT THE RELATIONSHIPS AMONG THOSE COLUMN NAMES. I do know some obvious relationships and am trying to find more in a statistical way. I have to pick the most possible correlatIve variables to do basic analysis. It is difficult if I don't know each variable's meaning. That's probably the reason why I posted such a vague question.
 
Joined
Mar 22, 2014
Messages
4,612
Likes
2,161
Location
Los Alamos, NM.
#7
TY. That's the point.
Maybe I should focus on the meaning of data in the same scope (prefix), such as IMUEX,IMU_ATTI ,OA, and try to find relationships among them.

Actually, the problem is that I WAS WONDERING ABOUT THE RELATIONSHIPS AMONG THOSE COLUMN NAMES. I do know some obvious relationships and am trying to find more in a statistical way. I have to pick the most possible correlatIve variables to do basic analysis. It is difficult if I don't know each variable's meaning. That's probably the reason why I posted such a vague question.
It's not so much that your question is vague - it's that it seems ill-posed. The data fields are either measured values from the various sensors, or quantities calculated from those values, such as time-integrals. I don't understand why you would use any kind of statistical approach to figure out how they are related.
 
Joined
Feb 10, 2019
Messages
8
Likes
0
Age
24
#8
I will think about your words carefully and consider using serveral variables to do a prediction.
The prediction probably is not included in the data log. That requires more related domain knowledge.:rolleyes:
 
Joined
Jun 24, 2015
Messages
1,321
Likes
490
Age
70
Location
Arroyo Grande, California
#10
@AMadCorgi the data descriptions you see at V3 .CSV column descriptions could certainly be better. The fact that several DJI platforms are included makes it a tedious, complicated business. And, I'm kinda lazy when it comes to documentation. To tell the truth I was kinda hoping that others would take an interest and help determine what some of these fields mean.

For some of the missing descriptions it's obvious what the description should be; e.g. BattInfo:BatTemp or BattInfo:Temp. So obvious that I could convince myself that a description isn't needed. For some of the fields like Motor:V_out:RBack I'm not confident enough to put forward a description. And, for several of the fields I have no idea what they are; e.g.IMUEX(imu#):us_p.

If you manage to determine what some of the descriptions should be please forward them.
 
Joined
Feb 10, 2019
Messages
8
Likes
0
Age
24
#11
Thank you all.
@sar104 To be honest, this is the first time I met a drone data log. So I would like to start building a neural network with some familiar and easy variables that I already knew. That sounds none of my interests' business. I will let you know if I have a new idea.
@BudWalker I will collect those fields that are hard to explain while doing the data process. Then list them here and let you know. It depends on how is my everything going. Hope it won't be long.
 
Joined
Jun 24, 2015
Messages
1,321
Likes
490
Age
70
Location
Arroyo Grande, California
#12
Thank you all.
@sar104 To be honest, this is the first time I met a drone data log. So I would like to start building a neural network with some familiar and easy variables that I already knew. That sounds none of my interests' business. I will let you know if I have a new idea.
@BudWalker I will collect those fields that are hard to explain while doing the data process. Then list them here and let you know. It depends on how is my everything going. Hope it won't be long.
You might want to look at the relationship between BattInfo:Ad_v and BattInfo:pack_ve. Both appear to be total battery voltage but they aren't exactly the same.
1550243179514.png
 
Joined
Feb 10, 2019
Messages
8
Likes
0
Age
24
#13
@BudWalker I would say they might have different Freq(HZ) at the first glance. Unfortunately, their Freq(HZ) are same. So the problem turns into what it refers to exactly.

Now i decide to focus on BattInfo:Remaining% (regarded as dependent variable) and collect every possible variables (regarded as independent variables) at first.
 
Joined
Jun 24, 2015
Messages
1,321
Likes
490
Age
70
Location
Arroyo Grande, California
#14
@BudWalker I would say they might have different Freq(HZ) at the first glance. Unfortunately, their Freq(HZ) are same. So the problem turns into what it refers to exactly.

Now i decide to focus on BattInfo:Remaining% (regarded as dependent variable) and collect every possible variables (regarded as independent variables) at first.
As the name implies BattInfo:Remaining% is the percentage of battery remaining. For a fully charged battery it starts at 100% and decreases as the flight progresses.
1550538312546.png
 
Joined
Feb 10, 2019
Messages
8
Likes
0
Age
24
#15
As the name implies BattInfo:Remaining% is the percentage of battery remaining. For a fully charged battery it starts at 100% and decreases as the flight progresses.
View attachment 108701
Nice. It's likely that there is a negative linear relationshipi between BattInfo:Remaining% and time.

But in most cases we are dealing with a more complicated data log, which means time is no longer the only independent variable to consider. For instance, acceleration, velocity, extral load, wind, higher or lower voltage(current), temperature, etc. Each of them could be represented by one or more detectable variables. I will find more relative detectable variables from data log.

I would be appreciated if I could get more inspiration and help from you in the future. Besides, why the time record starts with a negative number?
 
Joined
Jun 24, 2015
Messages
1,321
Likes
490
Age
70
Location
Arroyo Grande, California
#16
Nice. It's likely that there is a negative linear relationshipi between BattInfo:Remaining% and time.

But in most cases we are dealing with a more complicated data log, which means time is no longer the only independent variable to consider. For instance, acceleration, velocity, extral load, wind, higher or lower voltage(current), temperature, etc. Each of them could be represented by one or more detectable variables. I will find more relative detectable variables from data log.

I would be appreciated if I could get more inspiration and help from you in the future. Besides, why the time record starts with a negative number?
The time starts at a negative value because the time axis is offset so that time = 0.0 corresponds to the launch. The .DAT starts recording when the battery is turned on. The time axis offset can be changed.
1550593178995.png
 

Forum statistics

Threads
132,338
Messages
1,371,803
Members
96,625
Latest member
P3PUSER1928